169 69 27MB
English Pages 312 [313] Year 2021
Software Engineering Perspectives in Computer Game Development
Software Engineering Perspectives in Computer Game Development Edited by
Kendra M. L. Cooper
First Edition published 2021 by CRC Press 6000 Broken Sound Parkway NW, Suite 300, Boca Raton, FL 33487-2742 and by CRC Press 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN CRC Press is an imprint of Taylor & Francis Group, LLC © 2021 Taylor & Francis Group, LLC Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, access www.copyright.com or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. For works that are not available on CCC please contact mpkbookspermissions@tandf. co.uk Trademark notice: Product or corporate names may be trademarks or registered trademarks and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Names: Cooper, Kendra M. L., editor. Title: Software engineering perspectives in computer game development / edited by Kendra M. L. Cooper. Description: First edition. | Boca Raton : CRC Press, 2021. | Includes bibliographical references and index. Identifiers: LCCN 2020057801 | ISBN 9781138503786 (hbk) | ISBN 9781315145549 (ebk) Subjects: LCSH: Computer games--Programming. | Software engineering. Classification: LCC QA76.76.C672 S647 2021 | DDC 794.8/1525--dc23 LC record available at https://lccn.loc.gov/2020057801 ISBN: 978-1-138-50378-6 (hbk) ISBN: 978-1-032-01164-6 (pbk) ISBN: 978-1-315-14554-9 (ebk) Typeset in Times by SPi Global, India
Contents Preface.......................................................................................................................vii Editor..........................................................................................................................ix Contributors...............................................................................................................xi Chapter 1 Introduction to Software Engineering Perspectives in Computer Game Development.............................................................. 1 Kendra M. L. Cooper
SECTION I E merging Research on Serious Games for Software Engineering Education Chapter 2 Case Studies and Practices in Local Game Jam Software Development Organization: A Software Engineering Perspective...... 37 Walt Scacchi Chapter 3 Developing Essential Software Engineering Competencies Supported by Digital Game-Based Learning...................................... 59 Jöran Pieper Chapter 4 Building Casual Games and APIs for Teaching Introductory Programming Concepts....................................................................... 91 Brian Chau, Rob Nash, Kelvin Sung, and Jason Pace Chapter 5 Evolution of a Model for the Evaluation of Games for Software Engineering Education................................................. 103 Giani Petri, Christiane Gresse von Wangenheim, and Adriano Ferreti Borgatto
v
viContents
SECTION II E merging Research on Software Engineering Methods for Game Development Chapter 6 Applied Game Design Didactics....................................................... 141 Micah M. Hrehovcsik Chapter 7 Learning with the Body: A Design Framework for Embodied Learning Games and Simulations..................................................... 161 Edward F. Melcer and Katherine Isbister Chapter 8 Using Ideation Cards for Designing Mixed Reality Games.............. 197 Richard Wetzel, Steve Benford, and Tom Rodden Chapter 9 From Role-Playing Game to Exergame: A Modification for The Elder Scrolls V: Skyrim......................................................... 237 Rahul Kumar, Jak Tan, and Paul Ralph Chapter 10 Survey on Software Architecture, Creativity, and Game Technology........................................................................................ 253 Alf Inge Wang and Njål Nordmark Chapter 11 Games and Software Engineering Topics: Recent Results and Future Research Opportunities................................................... 279 Kendra M. L. Cooper Index....................................................................................................................... 295
Preface The tremendous success of today’s large-scale, complex computer games relies on an extensive history of advances in research and development that spans numerous disciplines. The earliest computer games emerged over seven decades ago as creations from research-oriented environments that possessed rare and expensive computing systems. In the 1950s, game developers were individuals who had access to mainframe computers and possessed the skills to write assembly code using punch tape. This small community was located in isolated research environments where they developed and played relatively simple games like tic-tac-toe to explore research topics or support public demonstrations. From these early beginnings, computer games have become ubiquitous. Their transition into popular cultures began in the late 1960s with on-going introductions of home console, arcade, personal computer, portable console, mobile phone, virtual reality/augmented reality, and multiplatform streaming games. The community continues to grow and embraces stakeholders with broad interests spanning entertainment games, serious games, gamification, and related research topics. Game development has emerged as a distinct research field with its own specialized journals (e.g., ACM Computers in Entertainment and IEEE Transactions on Games), conferences (e.g., Foundations of Digital Games and Game Developers Conference), and online communities (e.g., Games and Software Engineering Research). Universities worldwide offer Bachelor, Master, and Doctorate degrees in game development. In practice, today’s large-scale entertainment games are increasingly complex and can take years to develop. They have demanding traditional quality of service requirements: availability, backward compatibility, multiple platform deployment, performance, reliability, scalability, security, usability, and so on. In addition, they have distinct user experience requirements to provide the player with a game that is engaging and fun. In order to produce compelling user experiences, a wide variety of game elements and genres are used. Game elements include avatars, challenges, economies, flow, haptic and audiovisual interactions, progression (e.g., levels), rewards (e.g., points, badges, and leaderboards), and enabling social interactions among players. Genres include action, adventure, massively multiplayer online, role-playing, simulation, sports, and strategy. Game development relies on teams of specialists including game designers, artists, software developers (e.g., computer scientists and software engineers), and managers. The team composition in practice reflects the interdisciplinary foundations of game development, which include the arts and humanities; behavioral sciences; business; engineering; and physical sciences and mathematics. The increasing complexity of game development continues to drive innovative research. This book focuses on emerging results that focus on game development issues from software engineering perspectives. It is an excellent reference for software engineers, developers, educators, and researchers. To capture a current snapshot, this collection includes contributions prepared by 21 authors from nine countries; these have been provided by leaders in the community. Overall, the results vii
viiiPreface
are organized into two broad categories: serious games for software engineering education and software engineering methods for game development. The nine core chapters of the collection include five chapters on topics in serious games for software engineering education and four chapters on architecture and design methods for game development. In the first and final chapters, the interdisciplinary nature of games, a brief history of game development, and a summary of the recent peerreviewed literature (i.e., 2015–2020) on game development from software engineering perspectives are presented. The literature summary includes 18 articles on games for software engineering education and 65 articles on software engineering methods for game development. The articles are at multiple levels of maturity and appear in workshops, conferences, and journals. Based on the contributions in the book, possible future research opportunities are identified for consideration in terms of highlevel questions. The wide breadth of the discipline provides researchers with ample opportunities to identify and pursue topics of interest. et ludos incipere! Kendra M. L. Cooper
Editor Kendra M. L. Cooper is an independent Scholar and Editor. Her research interests are in software and systems engineering (requirements engineering, architecture) and engineering education; these topics are explored within the context of game engineering. Dr. Cooper has an extensive publication, teaching, and service record. Current research topics include the modelling, analyses, and automated transformations of complex game systems using graph-based methods; the ontological foundations of serious educational games are also under investigation. She has co-edited the book Computer Games and Software Engineering published by CRC Press. Dr. Cooper has co-organized/organized ICSE Workshops on Games and Software Engineering (GAS 2012, GAS 2013, GAS 2015, and GAS 2016) and the CSEE&T Workshop on Essence in Education and Training (2020). She has served as a coeditor of three JSPE Special Issues on Software Engineering in Practice (2018, 2019, 2020). She received a PhD in Electrical and Computer Engineering from The University of British Columbia, Canada.
ix
Contributors Steve Benford University of Nottingham Nottingham, UK [email protected] Adriano Ferreti Borgatto Federal University of Santa Catarina Florianópolis, Brazil [email protected] Brian Chau University of Washington Bothell Bothell, Washington [email protected] Micah M. Hrehovcsik HKU University of the Arts Utrecht Utrecht, Netherlands [email protected] Katherine Isbister University of California, Santa Cruz Santa Cruz, California [email protected] Rahul Kumar University of Auckland Auckland, New Zealand [email protected]
Njål Nordmark Norwegian University of Science and Technology Trondheim, Norway [email protected] Jason Pace University of Washington Bothell Bothell, Washington [email protected] Giani Petri Federal University of Santa Maria Santa Maria, Brazil [email protected] Jöran Pieper Hochschule Stralsund—University of Applied Sciences, IACS Stralsund, Germany [email protected] Paul Ralph Dalhousie University Halifax, Canada [email protected] Tom Rodden University of Nottingham Nottingham, UK [email protected]
Edward F. Melcer University of California, Santa Cruz Santa Cruz, California [email protected]
Walt Scacchi University of California, Irvine Irvine, California [email protected]
Rob Nash University of Washington Bothell Bothell, Washington [email protected]
Kelvin Sung University of Washington Bothell Bothell, Washington [email protected]
xi
xiiContributors
Jak Tan University of Auckland Auckland, New Zealand [email protected]
Alf Inge Wang Norwegian University of Science and Technology Trondheim, Norway [email protected]
Christiane Gresse von Wangenheim Federal University of Santa Catarina Florianópolis, Brazil [email protected]
Richard Wetzel Lucerne University of Applied Sciences and Arts Lucerne, Switzerland [email protected]
1
Introduction to Software Engineering Perspectives in Computer Game Development Kendra M. L. Cooper Independent Scholar, Canada
CONTENTS 1.1 1.2
The Interdisciplinary Nature of Computer Games�������������������������������������������� 1 A Brief History of Computer Games���������������������������������������������������������������� 5 1.2.1 Early Research Environment Games����������������������������������������������������� 5 1.2.2 Games in the Popular Culture��������������������������������������������������������������� 7 1.2.2.1 Home Console Games������������������������������������������������������������ 7 1.2.2.2 Arcade Video Games�������������������������������������������������������������� 9 1.2.2.3 Personal Computer Games��������������������������������������������������� 10 1.2.2.4 Portable Console Games������������������������������������������������������ 12 1.2.2.5 Mobile Phone Games����������������������������������������������������������� 13 1.2.2.6 Virtual and Augmented Reality Games�������������������������������� 13 1.2.2.7 Multiplatform Streaming Games������������������������������������������ 15 1.3 Recent Advances in Games and Software Engineering Research������������������� 16 1.4 The Chapter Collection����������������������������������������������������������������������������������� 16 1.4.1 Games and Software Engineering Education�������������������������������������� 16 1.4.2 Software Engineering Game Design��������������������������������������������������� 22 1.5 Summary��������������������������������������������������������������������������������������������������������� 26 References���������������������������������������������������������������������������������������������������������������� 26
1.1 THE INTERDISCIPLINARY NATURE OF COMPUTER GAMES Today’s success of large-scale, complex computer games relies on an extensive history of advances in research and development spanning numerous disciplines. The earliest computer games emerged over 70 years ago as creations from research- oriented environments that possessed rare and expensive computing systems. In the 1950s, game developers were individuals who had access to mainframe computers 1
2
Software Engineering Perspectives in Computer Game Development
and possessed the skills to write assembly code using punch tape. This small community was located in isolated academic or industrial research environments where they developed relatively simple games like tic-tac-toe to explore research topics or support public demonstrations. They were the original code hackers [62]. In this era, the software and hardware development infrastructure we take for granted today did not exist. For example, development methodologies, high-level programming languages, integrated development environments, specialized libraries (e.g., artificial intelligence, graphics, and physics), large-scale components (e.g., database management systems, cloud services, and game engines), multiuser operating systems, data communication networks, code repositories (e.g., GitHub), and inexpensive consumer hardware were not available. From these early beginnings, computer games have become ubiquitous. The community continues to grow and embraces stakeholders with broad interests spanning entertainment games, serious games, gamification, and related research topics. Game development has emerged as a distinct research field with its own specialized journals (e.g., ACM Computers in Entertainment and IEEE Transactions on Games), conferences (e.g., Foundations of Digital Games and Game Developers Conference), and on-line communities (e.g., The LinkedIn Games and Software Engineering Research Group https://www.linkedin.com/groups/13944027). Universities worldwide offer Bachelor, Master, and Doctorate degrees in game development. There are games for entertainment, games with a purpose (serious games), and gamified business applications. Entertainment games are a thriving industry with over two billion players around the world with 120 billion in revenues; higher numbers for both are anticipated in the future. In 2020, a long list of releases by major studios has been announced including the much anticipated sequels to Doom, Half-life, Halo, The Last of Us, and Wasteland franchises; new Marvel games include the Avengers, Iron Man VR, and Spider-man 2. Remakes to Final Fantasy VII, Resident Evil 3, and System Shock have also been announced as well as updated versions of Spongebob square pants and the 13th version of Microsoft’s Flight Simulator. The recent release of two ninth-generation consoles (Sony’s PlayStation 5 and Microsoft’s Xbox Series X) continues to drive game development. The AAA launch title for Microsoft’s new console is Halo Infinite; Sony’s launch event for their new console includes announcements for Spider-Man: Miles Morales, Horizon: Zero Dawn, and Resident Evil 7 sequels. Beyond entertainment games, serious games are available for education and training purposes; games are also used as example applications in non-gaming research. Young players can learn fundamental skills in reading and mathematics with simpler games. Business students play games to learn about finance and risk management topics. Computer science and engineering students can learn programming and software engineering skills (e.g., architecture, design, requirements engineering, project management, and testing). Medical students can learn anatomy with virtual reality (VR) applications that utilize Microsoft’s HoloLens. Serious games are available for training individuals in numerous domains such as fire safety, infrastructure inspection, manufacturing, and healthcare. Serious healthcare-related games on exercise, healthy lifestyles, mental health, patient care, and physiotherapy abound. More recently, gamified applications that integrate game elements (e.g., points, badges,
Introduction to Software Engineering Perspectives
3
and leaderboards) have emerged. In business environments, gamification efforts centre on improving the productivity and satisfaction of their employees by providing more engaging applications. In educational environments, learning management systems are gamified to improve the students’ learning outcomes and satisfaction. Games are also utilized to validate non-gaming research as well as for student projects in capstone design courses in computer science and software engineering degree programs. In practice, today’s large-scale entertainment games are complex and can take years to develop. They have demanding traditional quality of service requirements: availability, backward compatibility (e.g., earlier platforms), multiple platform deployment, performance, reliability, scalability, security, usability, and so on. In addition, they have distinct user experience requirements to provide the player with a game that is engaging and fun. In order to produce compelling user experiences, a wide variety of game elements and genres are used. Game elements include avatars, challenges, economies, flow, haptic and audiovisual interactions, progression (e.g., levels), rewards (e.g., points, badges, and leaderboards), and enabling social interactions among players. Game genres include action, adventure, massively multiplayer online (MMO), role-playing, simulation, sports, and strategy. Game development relies on teams of specialists including game designers, artists, software developers (e.g., computer scientists and software engineers), and managers. The team composition in practice reflects the interdisciplinary foundations of game development (Figure 1.1): arts and humanities; behavioural sciences; business; engineering; and physical sciences and mathematics. The more creative aspects of game development rely heavily on the arts, humanities, and behavioural sciences. From the arts, games utilize digital media (e.g., 2D/3D modelling), performing arts (e.g., music composition), visual arts (e.g., scene composition), and literature (e.g., storytelling via characters, narrative, plot, and setting). From the humanities, games invoke ethical questions addressed in philosophy, such as issues around game addiction. From the behavioural sciences, games rely on contributions in the disciplines of anthropology, sociology, and psychology. Anthropology supports game development in terms of considering the social and cultural norms of their target audiences. Sociology supports multiplayer game development in terms of the interactions of players in communities, both in real life and virtual. Psychology addresses topics around the gameplay experience: motivation, engagement, retention, and reward systems. For serious educational games, psychology encompasses learning theory which provides the core pedagogical foundation for these games. Game development also draws heavily from the physical sciences, mathematics, and engineering disciplines. Within the physical sciences, game research and development utilize results from numerous specialized subfields in computer science that began to emerge in the 1960s, including artificial intelligence, distributed/embedded/ information/real-time systems, human computer interactions, graphics, networking, operating systems, and programming languages. Today, these are well-established fields that remain active research areas; game developers continue to rely on these communities for innovative solutions. For example, advances in cloud computing research and development were embodied in an early cloud gaming streaming service technology, G-cluster; this service was demonstrated at the Electronic
4
Software Engineering Perspectives in Computer Game Development Industrial Design Software Engineering Systems Engineering…
Engineering
Digital Arts Literature Performing Arts Philosophy Arts Visual Arts…
Computer Science Dynamical Systems Physics Probability & Statistics…
Physical Sciences & Mathematics
& Humanities
Game Development Anthropology Communication Psychology Sociology…
Additional Fields…
Behavioral Sciences Business
Serious games on Architecture Business Education Engineering History…
Administration, Management, and Operations Marketing…
FIGURE 1.1 The interdisciplinary nature of game development.
Entertainment Expo in 2000. These services continued to evolve; games such as Homefront were released just over a decade later on OnLive’s streaming service. Game streaming services were subsequently introduced by established organizations (e.g., Google, Microsoft, Nvidia, and Sony) and start-ups; Amazon announced the release of their game streaming service, Luna, for 2021. From the engineering disciplines, game developers relied on innovations provided by the software engineering research and development communities. The software engineering discipline emerged in the late 1960s to address the challenges of developing large-scale complex applications. The declaration of the software crisis in 1968 was a pivotal moment: the cost of developing software systems had exceeded the cost of developing hardware [80]. The new field evolved into specialized subfields including lifecycle processes (e.g., Waterfall, Agile, and DevOps), engineering activities (e.g., requirements engineering, multiple levels of design, construction, and testing), umbrella activities (configuration management, project management, traceability), reuse (components, patterns, product lines), formal methods, modelling, automation (code, testing), and so on. In recognition of the 50th anniversary of the field of software engineering, articles summarizing its history from different perspectives were reported (e.g., [20, 21, 42]). Research at the intersection of games and software engineering emerged as a new discipline over the last decade (e.g., [16, 29, 30, 31, 91, 92, 109]); it spans the majority of subfields within software engineering (refer to Section 1.2).
Introduction to Software Engineering Perspectives
5
Although other established engineering disciplines such as systems engineering and industrial design have the potential to offer valuable contributions for game development, they have received little attention. Systems engineering emerged in the late 1940s by the US Department of Defense to develop large-scale, complex projects such as missile-defence systems. Today, established resources are available that address the issues in developing systems (e.g., [53, 55]). They provide guidance for specifying, designing, constructing, and validating/verifying systems (or systems of systems) that involve comprehensive technology stacks (e.g., communication networks, data storage, hardware, operating systems, and software). Many of these development issues are encountered in today’s large-scale games: they include distributed, real-time, multiplayer, multiplatform game systems that dynamically collect and analyse game play data. However, systems engineering contributions receive little attention in the game development community. Established industrial design methodologies are also available that explicitly address users’ emotional or affective responses to products. For example, Kansei engineering, originally introduced in 1974, relates customers’ emotional responses (i.e., physical and psychological) to the properties and characteristics of products or services [78]. In other words, they are designed to intentionally induce specific feelings. Kansei engineering continues to receive attention in a variety of communities (e.g., [54, 73]), but not in game development.
1.2 A BRIEF HISTORY OF COMPUTER GAMES Over time, the research and development of games has driven and leveraged numerous generations of computing devices and technology infrastructures. Games have evolved from specialized, limited use software applications to the ubiquitous products we see today. The history of games has been presented from a number of perspectives (e.g., [38, 46, 47, 110]). Here, games are organized into two main categories to provide a brief overview. The first reflects the early history of game development in research environments. The second category considers the shift of games into popular cultures; it is organized into groups including home console, arcade, personal computer, portable console, mobile phone, virtual and augmented reality (VR/AR), and multiplatform streaming games (refer to Figure 1.2).
1.2.1 Early Research Environment Games Researchers created a series of early games in the 1950s–1960s that generally simulated real-world board games. These games were rare, and the earliest of them were created when the specialized fields of computer science and software engineering did not yet exist. It was years before lower cost computing devices were available; the cost and complexity of developing software had not yet risen to the point where it was identified as a crisis. Games were developed as research projects to entertain the public or explore research problems in programming, human computer interactions, and computer algorithms. For example, a custom-built computer arcade game called Bertie the Brain was created at the University of Toronto. It was installed at the Canadian National Exhibition in 1950, where a player competed against the
6
Software Engineering Perspectives in Computer Game Development
Early, rare computer games • developed in research environments • academic • industry • military • small number of games, developers, and players
Evolution to ubiquitous Multiplatform streaming games computer games • widespread development VR/AR games (e.g., HMD, CAVE) • massive number of games and players Mobile phone games (2G-5G) • embedded in popular cultures Portable console games (handheld) Personal computer games (stand alone, networked) Arcade video games (cabinet machines) Home console games (TV)
Research environment games (e.g., academic, industry, military)
1950s
1960s CS emerges
1970s
SE emerges
1980s
1990s
2000s
2010s
present
Games & SE emerges
FIGURE 1.2 The shift from developing rare computer games in research environments to ubiquitous games in popular cultures began in the late 1960s with home consoles and arcade video games.
computer in a demonstration game of tic-tac-toe. Another computer called Nimrod was custom built by the company Ferranti; a player competed with the computer in a Nim demonstration game at the Festival of Britain in 1951. A single player version of tic-tac-toe called OXO was a game developed at the University of Cambridge in 1952. It ran on a computer called the Electronic Delay Storage Automatic Calculator (EDSAC). One of these machines was built by the university to serve their computing needs. The game program, based on a set of 14 available opcodes, was input using a punch tape and stored; hardware buttons, an analog telephone dial, and a green cathode ray tube (CRT) display provided the user interface devices for the game. The OXO game was developed as a research vehicle for investigating computer human interactions. A two-player game was developed called Tennis for Two in 1958. It was a demonstration game running on a commercial Donner Model 30 computer for the Brookhaven National Laboratory’s annual public exhibition. Spacewar! was another two-player, pioneering game. It was created for entertainment purposes and had a significant influence on the gaming community: it inspired many subsequent games. It was developed in 1962 at the Massachusetts Institute of Technology and ran on a commercial Digital Equipment Corporation PDP-11 minicomputer. The Spacewar! program, implemented using a set of 32 available opcodes, was input using a punch tape and stored in the minicomputer. The user interface devices consisted of a CRT display and a control box; the box had two knobs and a button. As lower cost hardware was introduced, the machines were adopted into more environments and eventually networked in research projects. These projects included the Programmed Logic for Automatic Teaching Operations (PLATO) in 1960,
Introduction to Software Engineering Perspectives
7
ARPANet in 1969, CYCLADES in 1971, NSFNet in 1979, BITNET in 1981, and the Internet in 1989. The first version of the PLATO system was pioneered at the University of Illinois at Urbana-Champaign; this research and development project was focused on supporting high quality online education [19]. This system evolved over four versions (i.e., PLATO I, II, III, and IV). PLATO I operated on an ILLIAC computer. It used a television set for the display and a special keyboard for one user to navigate the system’s function menus. PLATO II was available the following year, and it supported two users. The system was redesigned for PLATO III in 1964. This version operated on a Control Data Corporation 1604 computer and supported 20 terminals. The only remote PLATO III terminal was located near the state capitol in Springfield, Illinois, at Springfield High School. It was connected to the PLATO III system by a video connection and a separate dedicated line for keyboard data. The final version, PLATO IV, was introduced in 1972. It expanded into an internationally networked system of 10 sites that supported thousands of users and remained in service until 2006. A core contribution of the PLATO project was its role in developing an online community of users, which included game developers and players. The system was used to develop and share thousands of games in numerous genres including dungeon, space combat, simulations (e.g., flight and tank), maze, and card games. For example, networked games for 32 players developed for PLATO IV were Empire and Spasim. Empire began as a course project in 1973; it was a shooter-style game. Spasim was a space flight simulation and a first-person space shooter game that was released in 1974.
1.2.2 Games in the Popular Culture 1.2.2.1 Home Console Games The first home game console was the Brown Box prototype, which was released in 1967. It had six built-in games (chase-games, handball, light-gun, ping-pong, tennis, and volleyball). Magnavox licensed the prototype, and the first console appeared on the market in 1972 called the Odyssey; it had 28 games. Rather than having to go to an arcade, people could now play the popular arcade game PONG at home using a television as the display, albeit without sound. The initial console offerings were quickly expanded by Magnavox, Coleco, Atari, and Nintendo; these consoles played built-in, relatively simple games such as PONG, Hockey!, and Smash. In the late 1970s, numerous programmable ROM cartridge-based consoles emerged as the second generation. They allowed players to add new games developed for the proprietary consoles. Among these cartridge-based consoles was the Fairchild Channel F, which offered superior graphics, and the Atari 2600; both used eight-bit microprocessors. The Atari 2600 console dominated the market with the blockbuster game Space Invaders. To improve their development efforts for the console, Atari encapsulated reusable functionality to access the console’s display; this module was called the kernal. The commercial success of the consoles created opportunities for new studios to develop competing games that were less expensive, of variable quality, and widely available. In this changing business environment, Atari released two much anticipated titles: E.T. The Extra Terrestrial and PAC-Man. They were unsuccessful,
8
Software Engineering Perspectives in Computer Game Development
which contributed to The Crash of the video game market. As part of the major consolidation that followed, it took Atari 4 years to recover in North America. In the meantime, the popularity of consoles in Japan remained high. Within 1 month in 1983, three new home consoles were released in Japan: Microsoft Japan’s MSX hybrid computer-console system, Nintendo Entertainment System (Famicom), and Sega’s SG-1000. These consoles were popular in Japan; within a year, Ninetendo’s console dominated the market. Over the next decade, the intense competition in game consoles, known as the Console Wars [47], occurred involving Nintendo, Sega (USA and Japan), and later Sony. Nintendo offered Mario Brothers games and a reserved version of Mortal Kombat with a traditional marketing campaign. These were challenged by Sega’s irreverent Sonic the Hedgehog and an uncensored version of Mortal Kombat (complete with red-coloured blood and the option to make the game even bloodier) and an aggressive marketing campaign. After unsuccessfully negotiating a collaboration with Sega, Sony launched its own PlayStation console in 1994. Numerous generations of consoles have been offered by Microsoft, Nintendo, Sega, and Sony; the currently available consoles are considered the eighth generation. As an example, a brief summary of Nintendo’s home console product evolution is presented in Table 1.1. The table summarizes characteristics including the product name, generation, year, and technologies used. TABLE 1.1 Evolution of Nintendo’s Home Consoles Colour TV-Game. First-generation console, released in 1977. Only released in Japan (five versions). Each of the consoles contained a small number of built-in games and a controller. No microprocessor chips were used. Nintendo did not release a second-generation home console. Second-generation consoles deployed games in cartridges, used of microprocessor chips, provided single screen graphics (i.e., flip-screen). Nintendo Entertainment System Third-generation console, released in 1983. Embraced third party game development. Provided scrolling graphics. Customized chips: 8-bit Ricoh 6502 CPU, 8-bit Ricoh 2A03 GPU. Games deployed on ROM cartridges. Super Nintendo Entertainment System Fourth-generation console, released in 1990. Supported game play with the Interactor haptic feedback vest (1994). Customized chips: 16-bit Ricoh 5A22 CPU, 16-bit Ricoh PPU. Games deployed on ROM cartridges. Virtual Boy Fifth-generation console, released in 1995. First VR console with stereoscopic 3D graphics. The player uses the console like an HMD. Customized chip: 32-bit NEC V810 CPU. Games deployed on ROM cartridges. This product was quickly discontinued.
(Continued)
Introduction to Software Engineering Perspectives
9
TABLE 1.1 (Continued) Evolution of Nintendo’s Home Consoles Nintendo 64 Fifth-generation console, released in 1996. 3D graphics. Chips: 64-bit NEC VR4300 CPU, 64-bit Reality GPU. Connect up to four controllers (cables). Games deployed on ROM cartridges. Game cube Sixth-generation console, released in 2001. Chips: 32-bit IBM Gekko CPU, ATI Fliper GPU. External Ethernet IEEE 802.3 LAN and dial-up modem adapters offered. Games deployed on Game Cube Optical Discs (proprietary). Wii Seventh-generation console, released in 2006. Chips: 32-bit IBM Broadway CPU, ATI Hollywood GPU. Wii remote and Nanchuck devices, wireless Bluetooth, and infrared signals. Wi-Fi IEEE 802.11b/g for online gameplay at the Nintendo Wi-Fi Connection site. External Ethernet 802.2 LAN adapter available. Games deployed on Wii Optical Discs (proprietary). Wi U Eighth-generation console, released in 2012. Direct successor to the Wii console. Chips: 32-bit IBM Espresso CPU (quad core), AMD Radeon GPU (HD graphics) External Ethernet IEEE 802.2 LAN adapter available. Wii remote and Nanchuck devices, wireless Bluetooth and infrared signals. Wi-fi IEEE 802.11b/g/n for online gameplay on the Nintendo Network site. Games deployed on Wii U Optical Discs (proprietary). Nintendo Switch. Eighth-generation console, released in 2017. Convertible (home console/hand held). Customized SoC: NVIDIA Tegra X1 (64-bit, four ARM Cortex A57 CPU cores, 256 CUDA GPU cores). External Ethernet IEEE 802.2 LAN adapter available. Wi-fi IEEE 802.11b/g/n/ac for online gameplay at the Nintendo Switch site. Games and software are available on ROM cartridges and digital distribution. Nintendo Switch Pro. Ninth generation, announced for 2021. Successor to the Nintendo Switch.
1.2.2.2 Arcade Video Games The first two coin-operated arcade video games emerged in 1971: Galaxy Game and Computer Space. Galaxy Game was a prototype built at Stanford University using a PDP-11 minicomputer. The Computer Space game was a version of the pioneering game Spacewar! It was released by SyZyGy Engineering, the precursor of Atari, Inc. Soon after, the commercially successful arcade game PONG was released by Atari in 1972. This was followed by games that utilized innovative technologies and attracted a growing community of players. Game arcades flourished: they provided places for people to meet, play games, and have fun with friends. Popular games were released
10
Software Engineering Perspectives in Computer Game Development
in 1974–1977 including Taito’s Speed Race, which used scrolling sprites and a driving wheel; Midway MFG’s Gun Fight, the first arcade to use a microprocessor – the Intel 8080; Sega’s Moto-Cross, which provided haptic feedback; and Cinematronics’ Space Wars, the first arcade game that used vector graphics. The first blockbuster arcade game was Taito’s Space Invaders in 1978. In the next 3 years, iconic games including Atari’s Asteroids, Namco’s PAC-MAN/Ms.PAC-MAN, Nintendo’s Donkey Kong, and Williams Electronics’ Defender appeared. By 1982, there were 13,000 arcades in the USA, and each of the most popular machines brought in approximately $400.00 a week in quarters [94]. The trend turned abruptly a year later in North America, when the revenue of the game industry as a whole dropped drastically. This event, known as The Crash, impacted arcade games; however, they persisted. Some of the most popular arcade games of all time were released in the decade to follow, including Street Fighter/Street Fighter II/Street Fighter Champion Edition (Capcom, 1987, 1991, 1992), Mortal Kombat/Mortal Kombat II (Midway, 1992, 1993), and NBA Jam (Midway, 1993). Video arcades are rare today, but a nostalgic appeal for the games remains. Arcade games for the home market continue to emerge as either ¾ or full-sized cabinets. For example, full sized cabinet games were revealed at the Electronic Entertainment Expo 2019 event, including Arcade1UP’s Konami’s Teenage Mutant Ninja Turtles Home Arcade Game, Capcom’s Marvel Super Heroes Home Arcade Game, and an Atari Star Wars machine. 1.2.2.3 Personal Computer Games The growing popularity of personal computers in the late 1970s was embraced by the gaming community. The early offerings included the Commodore PET (1977), Tandy TRS-80 Model I (1977), Apple II (1977), and Atari 400/800 (1979) systems. IBM entered the personal computer market in 1981 with the 5150 model; a game friendlier home computer (graphics, sound, and price) called the PCjr was introduced in 1983. Although primarily for business applications and education, games occupied a prominent position in the software libraries of Apple, Atari, Commodore, IBM, Radio Shack, Texas Instruments, and other manufacturers. The game genres were diverse. Players purchased adventure, card, puzzle, simulation, and role playing games. For example, the first release of Microsoft’s Flight Simulator was offered in 1982 (the latest offering, version 13, was released in 2020). The first editions of very successful franchises were also released in the early 1980s. Origin Systems’ role playing game Ultima I The First Age of Darkness (1981) was released for the Apple II. This extremely popular game went on to offer nine sequels, game collections, and spin offs over the next three decades for multiple platforms. Sierra On-Line Systems’ adventure game King’s Quest (1983) was originally developed for the PCjr. Over the next 15 years, eight additional games in this franchise, as well as collections, were released for multiple platforms. In 1984, the game Tetris was created [3]. Originally programmed as a test program for the Electronika 60, it was ported and released for the IBM personal computer in 1986. Since then, Tetris and its variations were released on almost every platform; it’s been ranked as the best video game of all times [43]. Game development for personal computers expanded rapidly in the 1990s; new genres emerged including real-time strategy adventure (e.g., Dune), puzzle (e.g., SimCity), and first-person shooter (e.g., Doom) games. An early game engine to support the rapid, cost effective development of 2D games was available in 1993, IDs Doom engine. Game engines provided collections of
Introduction to Software Engineering Perspectives
11
reusable components (e.g., animation, artificial intelligence, collision detection, graphical user interfaces, inputs, loading, and physics). Rather than creating games from scratch, developers were able to focus on the unique assets and rules for their games. Today, game engines are widely available (e.g., Amazon Lumberyard, Blender, CryEngine, Unity, and Unreal Engine) and support multiplatform development. Beyond playing games on standalone personal computers, early advances in proprietary networking technologies provided new online gameplay opportunities. The Hayes 80-103a modem board was introduced in 1977, which utilized the S-100 personal computer bus architecture. An Apple II modem was introduced by Hayes the following year. These modems supported remote dial-up access via phone lines (i.e., using circuit connections) to bulletin board systems and online services. An early computer bulletin board system was called CBBS; it was a local bulletin board system available in 1978. Bulletin board systems flourished. The early systems were established and administered by hobbyists. They attracted computing enthusiasts primarily within their local telephone area codes. Users made free local calls via dial-up modems, as opposed to calling systems farther away that incurred long-distance phone charges. The bulletin board systems hosted Door games [40]. These games reflected the two steps players used to access games. Firstly, players connected to a bulletin board system. Secondly, they accessed a game hosted on the system. As an example, the Door game Trade War available in 1984 was a multiplayer, asynchronous turn play game on the Nochange bulletin board system. These bulletin board systems and Door games were very popular from the early 1980s to the mid-1990s. During this timeframe, proprietary online services were also launched to meet consumer demand (e.g., CompuServe, America on Line, and Genie). For example, CompuServe offered online services to the public in 1979. It went on to great success in the 1980s and remained a major influencer through the mid-1990s. CompuServe offered a catalogue of online games across adventure, dungeon, war, and board game genres. The catalogue included the very popular Island of Kesmai (1982) and Megawars III (1984). Users paid to play the games at hourly rates. Beyond playing the games, the community members were able to interact and discuss the games. The online services provided a space for gamers, game journalists, and game developers to share information. Recognizing the need to move from proprietary networks to open standard-based solutions, organizations were established for the Internet and subsequently for the World Wide Web (WWW). The earliest organization was the Internet Configuration Control Board; it was created by the ARPANet program in 1979. This board was reorganized and renamed over time. It is now called the Internet Architecture Board. The board first captured the state of standardization in 1988 for the Internet protocols in RFC 1083 [52]. At that time, the standards included the foundational RFCs for IP, ICMP, UDP, TCP, DNS, TELNET, FTP, and so on. Today, the board includes the Internet Engineering Task Force (IETF) that continues to oversee the development of Internet standards. An example of gameplay shifting from proprietary to open standard-based networks is the online game Legends of Future Past. It was initially released in 1992 on the Digital Dream Networks proprietary service; players accessed the game via a dial-up connection. The game was subsequently ported from the proprietary network to the Internet. Building on the Internet standards and technology, the WWW was created at the European Organization for Nuclear Research (CERN) in 1989 to support automated information sharing between scientists at universities and institutes. The first WWW
12
Software Engineering Perspectives in Computer Game Development
server and browser emerged the following year. The World Wide Web Consortium (W3C), established in 1994, was created to develop and maintain open standards for the WWW. MMO games took advantage of the new technology based on open standards. The real-time strategy game Earth: 2025 was an MMO browser game released in 1996; the MMO first-person shooter game Quake was released the same year. The real-time strategy games Age of Empires and StarCraft were released in 1997 and 1998, respectively. Advances in local area network (LAN) technology also provided new gameplay opportunities. For example, the AppleTalk LAN was available in 1985; network versions of games (e.g., Strategic Conquest in 1986 and Spectre in 1991) were released for the Macintosh. Proprietary LAN technology evolved into products based on open standards. The IEEE Ethernet LAN standard [51] enabled the low cost, mass production of network interface cards for IBM compatible personal computers; and wireless standards followed providing players with untethered gameplay experiences (e.g., [50]). Multiplayer, networked games such as Pathway to Darkness and DOOM launched in 1993; they initiated the LAN Party phenomenon. Game enthusiasts collected their personal computer equipment (desktop machines, monitors, and cables), gathered together, and set up a LAN to play their favourite games. These private LAN parties, very popular from the mid-1990s–2000s, have evolved. Today, players also gather virtually via social media platforms and large-scale commercially sponsored parties are organized. For example, the international DreamHack festival events draw several hundred thousands of participants including bring your own computer LAN parties. 1.2.2.4 Portable Console Games Handheld portable game consoles emerged in the mid-1970s that allowed players to enjoy their games on the go. These provided an alternative to playing games at home (console, personal computer) or in an arcade. In 1976, Mattel designed a game console the size of a calculator, using innovative displays based on light-emitting diode technology; the console played the game Auto Race. The next year, the Football game console was released. These electronic handheld games were very successful and attracted developers including Atari, Coleco, Parker Brothers, Milton Bradley, Entex, and Bandai. The first offering with interchangeable game cartridges was Milton Bradley’s Microvision in 1979. However, this console was short lived due to its fragile, small liquid-crystal display and limited selection of 12 games. Nintendo’s Game & Watch series of handheld consoles began its successful run in 1980, which lasted over a decade. A collection of 60 games was developed over a series of 11 consoles. The games included the iconic Donkey Kong (e.g., Donkey Kong, Donkey Kong Jr., and Donkey Kong II) and Mario franchises (e.g., Mario Bros., Mario’s Bombs Away, and Super Mario). Some of these games were later ported to the Game Boy, Game Boy Color, and Game Boy Advance consoles. In 1989, Nintendo popularized handheld gaming with the release of its 8-bit Game Boy video game device. For the North American launch, the titles included Super Mario Land, Alleyway, Baseball, Tetris, and Tennis. Over the next 25 years, handheld consoles were released by many organizations including NEC, Sega, Sony, and Nintendo for this growing community. Nintendo dominated the field and released a number of successful successors to the Game Boy, including the 8-bit Game Boy
Introduction to Software Engineering Perspectives
13
Color (1998) and the 32-bit Game Boy Advance series (Advanced 2001, SP 2003, Micro 2005). A collection of over 500 games was ultimately developed for the Game Boy Advance consoles. Nintendo’s next handheld consoles were in the 32-bit DS series (DS 2004, DS Lite 2006, DSi 2008, DSi XLs 2009, 3DS 2011, 2DS 2013); the Nintendo Switch/Switch Lite (2017) provided the option to operate either as a handheld or a traditional console. The backward compatibility of the new consoles in the DS series was a critical design constraint. Players who purchased a new console had access to an extensive game library; there are over 2000 games available today. The current handheld consoles are the eighth generation. 1.2.2.5 Mobile Phone Games The first games appeared on mobile cellular phones in 1994. They were preloaded, single player puzzle games on phones that utilized 2G cellular networks. Tetris was provided on the Hagenuk MT-2000 handset; Scramble was on the IBM Simon Personal Communicator smart phone. Three years later, the game Snake was provided on the Nokia 6110 mobile phone. The shift from playing only preloaded, single player games to being able to load new games for multiple players was a significant change. This advance relied on the Wireless Application Protocol (WAP) introduced in 1999 [108]. The protocol supported Internet access from mobile phones; WAP browsers emerged shortly after. Early on-line game sites included Starduck Central in 2001 (the predecessor to Impulse) and Valve Steam in 2003. The first Apple iPhone (IoS) was released in 2007 for 2G networks and the App Store followed the next year. This was the first on-line store dedicated to mobile apps. The game Texas Hold’em was the first game sold on the App Store in 2008. This mobile game supported up to nine players communicating via Wi-Fi. The HTC Dream phone (Android) for 3G networks and the on-line store Market were also available in 2008. The first commercially successful mobile game was Angry Birds, which was released in 2009. Over time, the technology and infrastructure for mobile phones continued to advance (e.g., 4G and 5G networks) and mobile games proliferated. This year, for example, the App Store’s most popular category was games. There were almost one million game apps available; these represent over 20% of the apps on their site. 1.2.2.6 Virtual and Augmented Reality Games 1.2.2.6.1 Virtual Reality Games The introduction of consumer VR devices began in the late 1980s. The devices for interactive gameplay expanded from traditional controllers, driving wheels, and joysticks to include devices spanning VR and AR environments that supported immersive gameplay [71]. Early devices for consoles and personal computers included a data glove, a wearable vest, and head-mounted displays (HMDs). The Mattel Power Glove was released in 1989; it was utilized in the e-sport Super Glove Ball VR game a year later for the Nintendo Entertainment System. Aura System’s Interactor Vest, a force feedback wearable device, was released and utilized in the fighter game Mortal Combat II for the Super Nintendo Console (SNES) and the SEGA Genesis versions in 1994. The Forte VFX-1 HMD was released and utilized in the first-person shooter game Descent in 1995 for personal computers running MS-DOS. The VFX-1 had
14
Software Engineering Perspectives in Computer Game Development
stereo headphones, stereoscopic displays, and three-axis head-tracking. This technology was acquired by Vuzix; it subsequently released two consumer HMDs: the VXF3D in 2000 and the iWear AV310 in 2009. More recently, Vuzix has focused on HMDs for commercial environments. Sony was another pioneer in consumer HMDs. It launched Glasstron in 1996, which was compatible with DOS personal computers and used in the game MechWarrior 2. Over a decade later, Sony launched a series of HMDs: HMZ-T1 in 2011 for the PlayStation 3; HMZ-T2 in 2012 for the Xbox 360 and PlayStation 3; and HMZ-T3W in 2013 with HDMI connectivity to consoles, mobile devices, and personal computers (wireless feature is optional). Sony also released the PlayStation VR in 2016 for the PlayStation 4 console; this product remained compatible with the PlayStation 5 console. Companies that joined the consumer HMD marketplace later include Oculus VR (a division of Facebook) and DreamWorld. Oculus VR released the Oculus Rift and Oculus Quest product lines. Their first product was the Rift CV1 consumer headset that was released in 2016; an upgraded version was available 3 years later. DreamWorld introduced the Dream Glass 4K in 2019. It provided state-of-the-art features including a 90° field of vision, 4K resolution, extensive play time, and compatibility with phones and consoles at an affordable price for home use. Beyond consoles and personal computers, VR was utilized in arcade games and virtual world environments. The Virtuality Group created VR arcade machines in 1991 that boasted short response time and provided gamers with stereoscopic vision devices, game controllers, and the ability to cooperate in multiplayer games. Nine games were developed for the machines, and the most popular one was the firstperson shooter Dactyl Nightmare. Cave Automatic Virtual Environments (CAVEs) emerged to support VR applications. The game CAVE Quake II was used to demonstrate the VisBox CAVE in 2001. CAVEs are immersive VR environments that utilize fixed projectors for three to six walls of a room-sized cube. They have a number of advantages for users: reducing the burden of carrying or wearing devices (e.g., HMDs), providing a shared experience for multiple people, and reducing the isolation of people from the real world. Currently, however, the CAVEs remain expensive to install which limits their adoption by consumers. 1.2.2.6.2 Augmented Reality Games Augmented reality (AR) games, providing a mix of a virtual world and elements from the real world, initially emerged for consoles. The game EyeToy was released in 2003 for the PlayStation 2 console. It used a TV screen for the display and a USB camera. For the first time, players were able to see themselves on-screen interacting with virtual objects in real time. The game Eye of Judgment was released in 2007 for the PlayStation 3 console. This console provided enhanced image processing capabilities using motion and colour detection in the PlayStation eye system. The game allowed two players to compete in online matches with a hybrid of paper and electronic collectible cards. AR games for handheld consoles soon followed. The game Ghostwire: Link to the Paranormal was an AR game first developed for the Nintendo DSi; this game was released in 2010. The game made use of the DSi’s camera and microphone so the player hunted for ghosts in their actual, physical environment. Once the ghosts were
Introduction to Software Engineering Perspectives
15
detected, the player tracked down objects and used them to appease the ghosts. The Nintendo 3DS handheld console was released in 2011 with a collection of six cards (five classic characters and a question mark) and two pre-installed AR games: Archery and Star Pics. The cards were used by placing them on a table and focusing the 3DS’ camera on them. The Nintendo Switch, a convertible console, was released in 2017. AR games were announced including the game Mario Kart Live: Home Circuit. This game uses a camera located on a remote control car and the Nintendo Switch console’s screen; its expected release date is in 2021. For mobile phones, early AR games included ARCade that allowed players to chase and consume virtual yellow dots like the Pac-Man game. It was developed as a tailored Layer browser in 2010 to celebrate the 30th anniversary of Pac-Man. The mobile AR game Pokémon GO became a global cultural phenomenon in 2016. At its peak, it attracted 28.5 million daily users. Players encountered, caught, and trained Pokémon creatures in real-world settings; it utilized the global position systems provided in the mobile devices. 1.2.2.7 Multiplatform Streaming Games Multiplatform game streaming services were offered over the last two decades that allowed players to run computationally intensive games (developed for specific platforms) on a variety of personal computers, consoles, and mobile devices. For example, rather than purchasing a dedicated home console or an expensive high-end gaming computer, games were accessible via modestly priced devices. Much of the processing in these services was allocated to server clusters to maintain the quality of the graphics and gameplay. Thin clients were used on the devices to access the services. The services evolved steadily in several directions: providing fundamental technology, providing subscription services for playing games users own, providing subscription services to provide access to games they don’t own. Here, examples of game streaming services available over time are briefly presented. A pioneering service was OnLive, which focused on providing subscriptions to game players. It was launched in 2003 and allowed subscribers to rent games without installing them on their own device. Games were delivered to OnLive’s client software as a streaming video that was rendered by the service’s servers. Gameplay was enabled through thin clients for personal computers, mobile devices, as well as smart TVs and a dedicated video game console-styled device called the OnLive Game System. Sony Interactive Entertainment acquired the patents for the OnLive technology in 2015, and the service was discontinued. G-cluster initially focused on providing streaming services to telecommunication companies. Their first release was in 2005 for the Cyprus Telecom Authority. Over the next decade, additional roll-outs with major telecommunication companies and content acquisition deal with studios (e.g., Electronic Arts, Gameloft, Sega, and Ubisoft) made the G-cluster technology a top provider. Gaikai focused on developing cloud-based game streaming technology; it was launched in 2011. A year later, the company was acquired by Sony Interactive Entertainment. The technology was adopted in the PlayStation Now game streaming service, which was released in 2014. The Sony service provided players access to over 300 games via their personal computer
16
Software Engineering Perspectives in Computer Game Development
(Windows). Shadow by Blade focused on providing players with remote access to a powerful personal computer (Windows) to play games they own. This service was released in 2016. Players accessed the services from a variety of devices including personal computers (Windows, macOS) and Android devices (smartphones, tablets) for a lag-free gaming experience. Players installed their games, preferences, and account information on a dedicated Windows 10 machine; everything is retained between sessions just like a local machine. Three game streaming services were launched in 2020: Nvidia Geforce, Google Stadia, and Microsoft Project xCloud. Nvidia GeForce focused on users playing games from their Steam library on personal computers (Windows, macOS), Android devices, or an Nvidia Shield TV. Google Stadia focused on providing access to console and personal computer games they don’t own via phones, Chrome browser tabs, or TVs. Microsoft Project xCloud focused on users accessing games they host; it provided 150 games at the launch. Players can currently use Android smartphones and tablets to play games; in the future the xBox one console streaming is planned.
1.3 RECENT ADVANCES IN GAMES AND SOFTWARE ENGINEERING RESEARCH Today’s large scale, complex games present distinct development challenges. Like many software projects in other domains, games frequently miss their originally announced release dates or are cancelled. For example, Doom Eternal, the fifth main game in the series, has recently been released for personal computers (Windows), consoles (PlayStation 4, Xbox One), and streaming (Stadia) platforms; the Nintendo Switch console port is forthcoming. This release was 4 months late and did not meet all of its requirements. The multiplayer invasion mode is not included in the release, which lets a player join another player’s game disguised as a demon. To address the challenges of developing large-scale complex games, new and customized approaches have been proposed in the literature. Table 1.2 presents a highlevel overview of recent results that span software engineering and game development research topics. Since 2015, five groups of topics have received considerable attention in the peer-reviewed literature: analytics and metrics, automation, models, reuse, and user experience. The work has appeared in computer science, education, software engineering, and game development publications. In addition, games for software engineering education and training have also recently been proposed. These include games on requirements engineering, architecture and design, modelling languages, code quality (testing, refactoring), lifecycle processes, and project management. The table is intended to identify the recent results in the games and software engineering research communities. It may provide a useful foundation for a comprehensive literature survey on games and software engineering in the future.
1.4 THE CHAPTER COLLECTION 1.4.1 Games and Software Engineering Education An exploration of organizational features that characterize game competitions and related game development environments is presented by Walt Scacchi in Chapter 2,
Introduction to Software Engineering Perspectives
17
TABLE 1.2 Examples of Recent Results: Spanning Games and Software Engineering Research Topics Software Engineering for Games Analytics and Metrics Game metrics [18, 27, 56, 83, 101] Game analytics [32, 57, 66, 77, 99] Automation PCG algorithms [2, 12, 24, 26, 65, 90, 95, 100] Testing [48, 49, 60, 82, 93, 98, 104, 112] Development Models and Methodologies Lifecycle process model [5, 6, 59, 61] Methodologies [13] Model-based development [37, 67, 113] Modelling languages [1, 9, 34, 91, 111] Reuse Frameworks [81, 105, 106, 107] Patterns [7, 41, 58, 63] Product lines [4, 35, 70, 96] Reference architectures [68, 74, 84, 97] User Experience Assessment and testing [17, 86, 88] Design [25, 79, 85] Requirements engineering [23, 33, 103] Serious Games for Software Engineering Engineering Activities Requirements [36, 44] Architecture and design [14, 76] Modelling languages (formal specification) [87] Code quality [11, 28, 39, 45, 72] (e.g., debugging, refactoring, and testing) Umbrella activities Lifecycle processes [8, 10, 15, 69] Project Management [22, 64, 75, 102]
followed by a qualitative analysis of five events the University of California, Irvine has been involved with. The organizational features focus on the setting of the event, which include academic settings (intra-mural, inter-mural, capstone projects), workshops, festivals, and additional variations; these events involve teams that typically work in parallel to make games. Intra-mural competitions involve teams from one school, whereas inter-mural teams involve teams from two or more schools. Capstone project courses utilize games as engaging problems that present teams with significant software engineering challenges. Game workshops provide a forum that offers a more structured event, including a curriculum and schedule of activities, to explore specific topics. The organizers involve subject matter experts to organize and present material on specific topics and lead interactive activities for the participants. Game showcase festivals provide a product-centric environment, where participants vie for recognition and awards for the creative games they have developed. These festivals
18
Software Engineering Perspectives in Computer Game Development
are like well-known film festivals (e.g., Sundance), in which the films are shown and critiqued, but little attention is paid to their production process. The five locally held events selected and analysed include two intra-mural game jams, a regional inter-mural game showcase, an intra-mural game capstone project course, and an international multiteam game development workshop focused on producing game modules for incorporation within a common game software architecture. Each event is described and comparatively analysed to present observations, lessons learned, and opportunities for how to organize and design game jams. The lessons learned and opportunities identified are summarized for the reader as nine items; they span topics of interest to researchers and educators. Recognizing the potential to utilize these competitions in empirical studies, there are fundamental questions around establishing rigorous, reusable methodologies that support the execution of repeated studies. The competitions offer a number of opportunities to explore additional possible organizations, in particular with respect to their suitability for investigating questions of interest to local, regional, national, or international populations. The numerous characteristics and trade-offs to consider for a project in a game development competition are also a rich source of research and education opportunities. For example, decisions on a game project competition span the purpose, duration, team formation, supervision, rewards, tools, and specific development skills to focus on. Such skills can include the development of large-scale courses (e.g., MOOCs), collaboration in distributed development, and use of game publishing platforms. While often focusing on programming aspects, development artefacts beyond the code can also be addressed including engineering (e.g., requirements, design, and testing) and project management. The characterization of the organizational features for game development events, as well as the lessons learned from the analysis of five events the University of California Irvine has been involved with, can facilitate playful learning and showcase opportunities in the future. Beyond the valuable educational opportunities, these events can also be organized to support the empirical evaluation of game software engineering research proposals. Digital-based learning games for software engineering education are the focus of the results presented by Jöran Pieper in Chapter 3. The need for students to develop transferrable, flexible software engineering competencies remains a challenging problem. These competencies are clusters of related knowledge, skills, and attitudes that can be applied to accomplish tasks in different contexts. The contexts span academic (e.g., course on software engineering, capstone project), and industry (e.g., internships, positions after graduation) environments. The use of serious games in software engineering education continues to receive the attention of researchers. The author explores the simulation games for software engineering that have been proposed in the literature; a survey is presented that includes 11 games evaluated using nine characteristics. The characteristics include gameplay (e.g., single-player or multiplayer) and subject domain (e.g., support for different SE processes/methods) specific criteria. The survey identifies an important gap in the literature: the games focus one well-established software process, rather than facilitating a perspective on the use of flexible SE practices; alternative processes are not embodied in the games. As
Introduction to Software Engineering Perspectives
19
a result, the currently available games do not focus on the development of transferrable, flexible competencies. To focus on a game that establishes flexible and transferrable SE competencies, the author selects the SEMAT Essence Kernel as a core element of the SE learning objectives. SEMAT Essence is an emerging OMG standard that provides a language for the unified description of SE practices and methods and a compact kernel. This kernel captures the essentials of SE endeavours in a process- and practice independent way; these essential elements can be selected and composed (i.e., flexible) to provide tailored SE methods that address the needs of a particular project (i.e., transferrable). The author develops two separate and complementary learning games, the Essence Kernel Puzzler and the Essence Simulation Game; in addition, a recommendation tool is also presented called the Essence Navigator. By separating the two games, their learning outcomes and game genres have been identified and set independently, rather than attempting to create a single game with a broad scope. In addition, they can be adopted as needed by instructors to support a course curriculum and schedule. The games are both established on social constructivist learning theory to provide a rigorous pedagogical foundation and the Revised Bloom’s taxonomy is used to characterize the learning objectives. The Essence Kernel Puzzler game addresses the introduction of concepts and vocabulary at lower competency levels, remember and understand in the knowledge dimension. Two genres are selected for this game: drill and practice; and puzzle games. This game prepares the student to play the second game, the Essence Simulation Game, which addresses more advanced learning objectives at higher competency levels, remember, understand, and apply in the knowledge dimension. It utilizes the features and characteristics of adventure and simulation games to impart the desired procedural knowledge and skills. The learning content is an integral part of the gameplay and is tightly linked to the SE domain. The design couples the mastery of the learning content with achieving success in the game; a high degree of reflection is embodied in this simulation game to provide the time needed to think and rethink decisions. These two games are complemented by the Essence Navigator, a supporting tool, connecting game experiences and realworld course project work of students. This tool has been integrated into the Essence Simulation Game to help monitor the progress of a virtual team’s project; it provides basic recommendations to the player on the next step to take. The navigation tool can also be used separately, for example, in a capstone project course, to assist teams. The proposed game-based learning approach, which includes two games and a navigation tool, is validated with a case study conducted with students. A questionnaire is used to collect data on the utility of the approach. The analysis of the data collected indicates the approach is promising; however, additional research is needed to assess the value of game mechanics including “time travel” (to provide opportunities to reflect upon earlier decisions, make alternative choices), rankings, and leaderboards. A collection of serious educational games targeting introductory programming concepts is under development by Brian Chau et al. in Chapter 4. The games within the collection are small, independent, and fun. Each game focusses on one or two
20
Software Engineering Perspectives in Computer Game Development
specific programming concepts (e.g., conditionals and arrays), which allows instructors to flexibly choose an ensemble of games to meet their needs. Each game is designed, built, and extensively play tested almost entirely by undergraduate students, ensuring they are fun and engaging for the target audience. For each game, an API is extracted which allows instructors and students to modify the game. Faculty can construct small and fun games to demonstrate concepts; students can exercise their own understanding and creativity by customizing the game and making it their own. The games are variations of popular casual genres; two completed games are Space Smasher (a variant of Super Breakout-style games) and Corrupted (a variant of Bubble Shooter games). The Smasher gameplay involves players removing blocks on the screen by bouncing a ball with a moveable paddle. Space Smasher introduces more interesting gameplay by adding customizable blocks that are capable of triggering events such as swapping blocks or enabling/disabling unbreakable blocks. Also included are more premium sets of graphics tiles and sounds than are typically found in this genre. The ball-block-paddle collision tests, the special event logic, and the iteration through all blocks present an excellent structured sandbox for teaching and playing with conditionals and loops. The Corrupted gameplay involves players launching a coloured tile into a larger group of tiles and attempting to make matches of three until all tiles are removed or until the tiles advance to meet the player at the opposite end of the screen. Bubble shooters also tend to be simple match-and-remove games with minimalist graphics. Corrupted recasts the game with an active automated opponent employing a variety of tricks to increase the challenge and intrigue, and the game itself has been given a distinct artistic style. The visual, spatial, and multidimensional aspects of the colour tiles present a rich domain of concepts for use in teaching 1D and 2D arrays. The research adopts four design goals for the API development. Firstly, the usability and simplicity of the APIS facilitate the creation of effective CS1/2 materials by users with no background in graphics or gaming. Secondly, users must be able to build simple applications quickly with minimum familiarity of the APIs; over time the users can gradually explore more the advanced functionality at their own pace. Thirdly, the APIs primarily support building effective teaching materials as vehicles for educational content delivery. Lastly, the APIs encapsulate the graphics and gaming functionality while selectively exposing concepts for educational purposes. Based on two completed games and their associated teaching materials, feedback from novice student programmers indicates that the games are engaging and the associated APIs are straightforward to use. The games are currently being field tested in CS1/2 classrooms. A new version of a model for systematically evaluating the quality of games for software engineering education is presented by Giani Petri et al. in Chapter 5. An existing model for the evaluation of educational games (MEEGA) has already been proposed in the literature; it has been widely adopted. MEEGA can be applied to evaluate educational games (digital and non-digital ones) in terms of motivation, user experience, and learning; it utilizes a standardized questionnaire to measure the reaction of students after playing the game. A thorough analysis of MEEGA is reported in this chapter, which reveals opportunities for its improvement related to the presence of overlapping theoretical concepts (motivation and user experience), in addition to some ambiguous phrasing in the questionnaire. The authors then propose an
Introduction to Software Engineering Perspectives
21
enhanced model to address these concerns called MEEGA+; it improves the reliability and validity of the model. This chapter reports on the design and evaluation of MEEGA+. The design consists of four steps. Firstly, the factors to be measured are defined to provide the foundation for the measurement instrument (questionnaire). This is achieved using the established Goal/Question/Metric approach, in which the high-level evaluation objective is defined and is systematically decomposed into specific factors for measurement. Secondly, the measurement of these factors is operationalized by decomposing them into measurement instrument items (i.e., questions). The specific learning objectives of a game are included in these items. In order to standardize the selected items, all of the items are transformed into positive statements. Thirdly, the response format for the items of the measurement instrument is defined (i.e., 5-point Likert scale). Lastly, an expert panel composed of four senior researchers with backgrounds in computing/SE and/or statistics is convened to analyse the clarity, relevance, consistency, and completeness of the measurement instrument items. Feedback from this panel has been used in the design of MEEGA+. The evaluation of the model focusses on three analysis questions related to evidence for internal consistency, how underlying factors influence the responses, and evidence for the convergent and discriminant validity of the measurement instrument. An extensive case study evaluation is reported on the evaluation of MEEGA+. It consists of 29 case studies, spanning 13 games (four digital, nine non-digital), responses from 589 students, and six different educational institutions. The results indicate the reliability of the MEEGA+ model as satisfactory; two quality factors have a significant role (player experience and usability). Thus, MEEGA+ provides a valuable contribution in the systematic development, improvement, and adoption of serious educational games for SE. An assemblage of applied game design methods for educators is presented by Micah Hrehovcsik in Chapter 6 to help address the current lack of a comprehensive design theory, best practices, and didactic approaches. Applied games are those that go beyond entertainment, for example, serious games or games with a purpose. The proposed collection relies on the Vitruvius principles. These principles are the foundation for ancient Roman architecture, which forms the triad of utilitas, firmitas, and venustas. Repurposed for applied game design, the principles encourage critical design awareness and offer a perspective on balancing a serious game’s utilitas (purpose), firmitas (sustainability), and venustas (gameplay experience). Utilitas, or purpose, is when a game fulfills its tactical purpose. Firmitas, or sustainability, is when the game is properly embedded in the context, available to players, has a syllabus designed around it, and aims to create a perceivable impact in the chosen domain. Venustas, or gameplay experience, is when the game provides a meaningful holistic experience (e.g., gameplay, graphics, and sound) for the player. The assemblage has been used to educate game designers at the HKU University of the Arts Utrecht about the design of applied games. It addresses three instructional goals: provide students with a means to critique serious games, provide a means to categorize games according to their design, rather than their domain or genre, and provide a tool which guides design choices, spurs design research, and supports the rationale for design choices. The assemblage consists of four components. The first component is a classification scheme for serious games, which has a collection of
22
Software Engineering Perspectives in Computer Game Development
tactical forms: transmissive, aggregative, adaptive, and collaborative. The transmissive form, for example, is designed to attract and engage players to ‘transmit’ skills, knowledge, rhetoric, therapy, etc., in serious games. The second component is a scope model, which supports the analysis of the content, context, and knowledge transfer by asking the classic who, what, where, when, why, and how questions. Once these questions have been answered the game designer has a quick-scan, or summary, that is used for subsequent steps in the design of the applied game in question. The third component is the use of epistemic cards as a tool for structuring the dissemination of design decision or understanding how design decisions are being made. There are six kinds of cards that define this epistemic framework including context, content, and transfer analysis; stakeholders; design; artefact; process; and project. These kinds of cards represent a collection of skills, identities, interests, understanding, and knowledge. Lastly, the game jam format provides an intensive practice-based learning experience concerning applied game design theories, codesign, best practices, and studio operations related to development. The format values putting students under time and peer pressure, while having them work autonomously in multidisciplinary teams.
1.4.2 Software Engineering Game Design A novel design framework for embodiment is presented by Edward Melcer and Katherine Isbiter in Chapter 7. Embodiment is an important concept which recognizes cognition as emergent and multifaceted, involving the mind, physical interactions, and social interactions in the environment. It has significant potential to enhance the design of educational games and simulations and has been investigated by diverse communities, including games. These investigations have resulted in the use of terms (e.g., embodiment, embodied cognition, and embodied interaction) with distinct definitions and operationalization, making the concept difficult to apply in practice. The authors propose a new framework to synthesize these results that supports the design for digital and non-digital systems. It is based on a careful selection and analysis of 90 articles that focus on embodiment and design descriptions, spanning 66 distinct games and simulations. The analysis results are used to establish the design framework taxonomy. The taxonomy is intended to make designers aware of design choices, rather than prescribing strict mappings of design choices into values, like sorting bins. It is organized into three levels: groups, dimensions, and possible values (i.e., design choices). At the highest level there are three groups: physical interactions, social interactions, and the world. The physical interactions group describes how learning is physically embodied in a system. This group has four dimensions, which are physicality, transformity, mapping, and correspondence. Within the physicality dimension, for example, one value is the augmented option, which refers to the use of an AR system as a design decision. The social interactions group describes how individuals socially interact, play, and coordinate within a system. This group has two dimensions: mode of play and coordination. For example, the mode of play dimension has three design options which are individual, collaborative, and competitive. The world group describes the context of the embodied learning and has a single dimension, the environment. This dimension has three values: physical, mixed, and virtual.
Introduction to Software Engineering Perspectives
23
The proposed taxonomy is evaluated using three example studies. The first study examines the ability to describe and categorize embodied games and simulations using the framework. The second and third studies apply the framework to identify problematic design spaces and design gaps, respectively. These studies indicate that the framework elucidates design decisions by providing a unifying foundation for the categorization and analyses of designs for existing games. The potential impact of the framework for designing new games and simulations is promising. The design of mixed reality games including pervasive games, location-based games, and AR games is explored by Richard Wetzel et al. in Chapter 8. These mixed reality games enrich the physical world with technology to create new and exciting possibilities for games. For example, a player’s location can be tracked with a GPS service in real time; arriving at a physical location can trigger changes in the game play. The new approach proposed by the authors is based on ideation cards, which are an established design technique. The physical cards are the size of traditional playing cards; each one focuses on a specific concept. The cards often provide a brief description of the concept and an indicative image. They may also provide a more detailed explanation or an example of its use. The mixed reality game card desk consists of Opportunity Cards, Question Cards, Challenge Cards, and Theme Cards. Opportunity Cards highlight game elements that may be adopted, such as tracking sensors. Question Cards probe a deeper exploration of design decisions, such as defining a design concept, or alternative scenarios. Challenge Cards present designers with issues that arise in mixed reality games, such as safety, adverse weather conditions, or congestion at a particular location. Theme Cards are not specific to mixed reality or games; their purpose is to provide additional inspiration for designers. The Theme Cards are realized using a collection of existing cards with images of abstract paintings. The final deck of the Mixed Reality Game Cards contains 93 unique cards (51 Opportunity Cards, 18 Question Cards, and 24 Challenge Cards). The cards are organized into nine categories to both make them more appealing and distinguishable. These categories are audio, gameplay, locations, management, physical, players, sensors, technology, and time. A lightweight design process is also introduced that consists of two activities. The first activity is idea generation, in which the goal is to use the cards to rapidly generate numerous ideas over several iterations. In each of the iterations, three Opportunity Cards and one Theme Card are randomly drawn to initiate discussions. Ultimately, one idea is selected to pursue. The second activity has four phases. The first is the idea development in which ideas are explored with Opportunity Cards. Secondly, the ideas are refined with Question Cards. Thirdly, the ideas are grounded with Challenge Cards. Lastly, the ideas are finalized by spatially organizing the cards and removing irrelevant ones. The authors report their experiences using the process and the card deck to develop a game for a non-profit organization. A novel design for an exergame is presented in Chapter 9 by Rahul Kumjar et al. The purpose of exergames is to motivate players to participate in physical exercise by providing enjoyable, immersive environments. The current designs of many exergames present some limitations as effectively combining immersive environments and intense activities at the same time is challenging; intense activity is considered
24
Software Engineering Perspectives in Computer Game Development
by some as the most beneficial kind of exercise. For example, playing a game and cycling on roadways at the same time can present safety concerns. Consequently, the design of an exergame that promotes exercise outside of play sessions is investigated by the authors. To accomplish this, a popular role playing game “The Elder Scrolls V: Skyrim” is redesigned as an exergame. The original game is multilevel, which is central to the players’ progression in the game. The player character has three attributes: health, stamina, and magicka. Health is how much damage a character can take before dying; stamina is the energy system used for running and performing special physical attacks; and magicka is the energy system for performing magic. It is a popular game available on multiple platforms, and it has an extensive community of users and collections of content (over 300 hours). A modification kit is built into the game, which is called the Creation Kit. This kit provides a graphical interface and allows players to readily create new quests, non-player characters, texture-packs, armour models, and so on. The kit uses a proprietary language, Papyrus, to support the modifications. In addition, the modding community for the game has produced resources and documentation for the kit and made them widely available on forums and Internet relay chat channels. In the modified game called the Skyrim Exergaming Mode, the player character levels up based on their real-world physical activity, rather than their in-game activity. The exergame design model weaves together physical exercise, quests, points, and a levelling system. Overall, there are four main steps in the exergame. In the first step, the player accepts a quest while playing the game. Later, the player engages in physical activities outside the game (e.g., fitness classes, stretching, and sport). In the second step, they log their activity using the exercise logging platform called Exercise. In the third step, the player opens the game; the game acquires the exercise data from Exercise to update the player’s points and levelling status. The player is notified of their updated levelling status and resumes playing the game in the fourth step. The design maps alternative types of physical activity as points into the character’s attributes. For example, weight training increases maximum health, cardio increases maximum stamina, and everything else (sports, stretching, and so on) increases the maximum magicka. The inherent progression of the game difficulty in the levelling design creates an incentive for players to continue exercising to make progress in the game. Chapter 10 by Alf Inge Wang and Njål Nordmark design, run, and analyse a twopart survey that explores a variety of perspectives on software engineering and game development. The research goals are to delve into how game developers think about and use software architecture in the development of games, how creative development processes are managed and supported, and the use of game technology. The authors define four research questions (refer to Table 1.3). In the first part of the study, a questionnaire consisting of 20 statements is prepared which are derived from the research questions. The respondents provide their responses using the Likert’s scale. In addition, the questionnaire provides a free text comment area for every statement. There are 13 subjects for the study; these are a mix of attendees at the Nordic booth at the Game Developer Conference in San Francisco and game developers contacted by email. The questionnaire has been answered both on paper and using web-forms created using SurveyMonkey. The results show that software
Introduction to Software Engineering Perspectives
25
architectures play a central role in game development; the main focus is on achieving software with good performance and high modifiability. The creative processes are supported using a wide variety of tools such as game engines, middleware, scripting languages, and dynamically loading assets. Feature-based teams have shifted over the last several years to using more game-specific engines, tools, and middleware in their development; these teams consist of both creative and technical professionals. TABLE 1.3 Questionnaire Summary Question No.
Questionnaire Statement
RQ1: What role does software architecture play in game development? 1
Design of software architecture is an important part of our game development process. 2 The main goal of our software architecture is performance. 3 Our game concept heavily influences the software architecture. 4 The creative team is included in the design of the software architecture. 5 Our existing software suite provides features aimed at helping the creative team do their job. 6 Our existing software architecture dictates the future game concepts we can develop. RQ2: How do game developers manage changes to the software architecture? 7 The creative team has to adopt their ideas to the existing game engine. 8 During development, the creative team can demand changes to the software architecture. 9 Who decides if change-requests from the creative team are implemented? 10 The technical team implements all features requested by the creative team. 11 It is easy to add new gameplay elements after the core of our game engine has been completed. 12 During development, the creative team has to use the tools and features already available. RQ3: How are the creative processes managed and supported in game development? 13 Our game engine supports dynamic loading of new content. 14 Our game engine has a scripting system the creative team can use to try out and implement new ideas. 15 The creative team is included in our development feedback loop (e.g., scrum meetings). 16 Our game engine allows rapid prototyping of new levels, scenarios, and NPCs/behaviour. RQ4: How has game development evolved in the last couple of years? 17 Today our company uses more 3rd party modules than 3 years ago. 18 It is easier to develop games today than it was 5 years ago. 19 Middleware is more important to our company today than 3 years ago. 20 Game development is more like ordinary software development today than 5 years ago.
26
Software Engineering Perspectives in Computer Game Development
After receiving the questionnaire responses from part one, a follow-up survey with eight open-ended questions is sent to six subjects; these individuals had participated in the first questionnaire and indicated they would provide more detailed answers. This part of the study has been conducted on the web only using SurveyMonkey. Summaries of the results are reported by the authors and are organized around three areas: the future of game engines; how the creative team affects the software architecture; and how companies are reasoning and making decisions on implementing changes. The respondents describe the future of game engines with four key quality attributes: multiplatform, quality of the features provided, simplicity of use, and completeness with respect to its integration into a diverse tool suite. The creative team indirectly affects the software architecture through requests made to the technical team; the main areas affected are related to how tools interact with the game. The respondents strongly favour the use of integrated feature-based teams and rapid prototyping to enable the interactions among the technical and creative team members. With respect to managing changes, a two-step process involving the assessment of the change in terms of the gameplay experience and the cost is revealed as a common pattern. In addition, the responses indicate the management, technical, and creative teams all providing input to the decision-making process for making changes; ultimately the management team makes the decision.
1.5 SUMMARY This chapter introduces a collection of recent research papers that provide on software engineering perspectives on computer game development. It begins with a bird’s eye view of game development as an interdisciplinary problem. It draws upon the arts and humanities, behavioural sciences, business, engineering, physical sciences, and mathematics. In addition, serious games can be developed for any domain. A brief history of computer games is presented, which is organized into two main categories: early research environment games and games for popular cultures. This followed by a snapshot of recent peer-reviewed publications focused on interdisciplinary work in games and software engineering. The nature of game development incurs significant complexity, which may, at least in part, be addressed by software engineering techniques. The following chapters present the core of this book, which is followed by a brief outlook on the future of research in computer games and software engineering.
REFERENCES [1] Abbadi, M. et al. (2015) Casanova: A Simple, High-Performance Language for Game Development. Vol. 9090. [Online]. Cham: Springer International Publishing. [2] Abuzuraiq, A. M. et al. (2019) ‘Taksim: A Constrained Graph Partitioning Framework for Procedural Content Generation’. In 2019 IEEE Conference on Games (CoG). [Online]. August 2019 IEEE. pp. 1–8. [3] Ackerman, D. 2016 The Tetris Effect: The Game that Hypnotized the World, PublicAffairs. [4] Åkesson, J., Nilsson, S., Krüger, J., & Berger, T. (2019). Migrating the Android ApoGames into an Annotation-Based Software Product Line. In Proceedings of the 23rd International Systems and Software Product Line Conference - Volume A (SPLC ’19). New York, NY, USA: Association for Computing Machinery, 103–107. doi: 10.1145/3336294.3342362
Introduction to Software Engineering Perspectives
27
[5] Aleem, S., Capretz, L. F., & Ahmed, F. (2016). Critical success factors to improve the game development process from a developer’s perspective. Journal of Computer Science and Technology, 31(5), 925–950. [6] Aleem, S., Capretz, L. F., & Ahmed, F. (2016). Game development software engineering process life cycle: A systematic review. Journal of Software Engineering Research and Development, 4(1), 6. [7] Allison, F., Carter, M., Gibbs, M., & Smith, W. (2018). Design patterns for voice interaction in games. In Proceedings of the 2018 Annual Symposium on Computer-Human Interaction in Play (CHI PLAY 18). Association for Computing Machinery, New York, NY, USA, 5–17. DOI:10.1145/3242671.3242712 [8] Ammons, B., & Bansal, S. K. (2017). Scrumify: A software game to introduce Agile Software development methods. Journal of Engineering Education Transformations. doi: 10.16920/jeet/2017/v0i0/111752 [9] Arias J., Marczak R., & Desainte-Catherine M. (2019) Timed automata for video games and interaction. In: Lee N. (eds) Encyclopedia of Computer Graphics and Games. Springer, Cham. doi: 10.1007/978-3-319-08234-9_298-1 [10] Aydan, U., Yilmaz, M., Clarke, P. M., & O’Connor, R. V. (2017). Teaching ISO/IEC 12207 software lifecycle processes: A serious game approach. Computer Standards & Interfaces, 54, 129–138. [11] Baars, S., & Meester, S. (2019, May). CodeArena: Inspecting and improving code quality metrics using minecraft. In 2019 IEEE/ACM International Conference on Technical Debt (TechDebt) (pp. 68–70). IEEE. [12] Baldwin, A., Dahlskog, S., Font, J.M., & Holmberg, J. (2017). Towards pattern-based mixed-initiative dungeon generation. In Proceedings of the 12th International Conference on the Foundations of Digital Games (FDG’17). New York, NY, USA: Association for Computing Machinery, Article 74, 1–10. doi: 10.1145/3102071.3110572 [13] Barnard, J., Huisman, M., & Drevin, G. R. (2018). The development of a systems development methodology for location-based games Computers in Entertainments, 16 (3), Article 1 (September 2018), 47. doi:10.1145/3236492 [14] Bartel, A. & Hagel, G. (2016, April). Gamifying the learning of design patterns in software engineering education. In 2016 IEEE Global Engineering Education Conference (EDUCON) (pp. 74–79). IEEE. [15] Baumann, A. (2020). Teaching software engineering methods with agile games. In Proceedings of the 2020 IEEE Global Engineering Education Conference (EDUCON), Porto, Portugal, pp. 1550–1553, doi: 10.1109/EDUCON45650.2020.9125129. [16] Bell, J., Cooper, K., Kaiser, G., & Sheth, S. 2012. Proceedings of the Second International Workshop on Games and Software Engineering: Realizing User Engagement with Game Engineering Techniques, GAS, Zurich, Switzerland, June 9, 2012. IEEE/ACM 2012, ISBN 978-1-4673-1768-9 [17] Bernhaupt, R. (2015). User experience evaluation methods in the games development life cycle. In Game User Experience Evaluation (pp. 1–8). Cham: Springer. [18] Birk, M. V., Lürig, C., & Mandryk, R. L. (2015). A metric for automatically flagging problem levels in games from prototype walkthrough data ACADEMICMINDTREK 2015 - Proceedings of the 19th International Academic Mindtrek Conference, pp. 33–40. [19] Bitzer, D., Braunfeld, P., & Lichtenberger, W. (1961). PLATO: An automatic teaching device. IRE Transactions on Education, 4(4), 157–161. [20] Booch, G., “The history of software engineering.” In IEEE Software, vol. 35, no. 5, pp. 108–114, September/October 2018, doi: 10.1109/MS.2018.3571234. [21] Broy, M. “Yesterday, today, and tomorrow: 50 years of software engineering.” In IEEE Software, vol. 35, no. 5, pp. 38–43, September/October 2018, doi: 10.1109/ MS.2018.290111138.
28
Software Engineering Perspectives in Computer Game Development
[22] Calderón, A., Ruiz, M., & O’Connor, R. V. (2017, September). ProDecAdmin: A game scenario design tool for software project management training. In European Conference on Software Process Improvement, Cham. pp. 241–248. Springer. [23] Callele, D., Dueck, P., Wnuk, K., & Hynninen, P. (2015). Experience requirements in video games definition and testability. In IEEE 23rd International Requirements Engineering Conference (RE), Ottawa, ON, 2015, pp. 324–333, doi: 10.1109/ RE.2015.7320449. [24] Campos, J. & Rieder, R. (2019). Procedural Content Generation using Artificial Intelligence for Unique Virtual Reality Game Experiences. In 2019 21st Symposium on Virtual and Augmented Reality (SVR), Rio de Janeiro, Brazil. pp. 147–151. [25] Canossa, A., Badler, J. B., El-Nasr, M. S., & Anderson, E. (2016). Eliciting emotions in design of games-a theory driven approach. In EMPIRE@ RecSys. Boston, MA, USA. pp. 34–42. [26] Capasso-Ballesteros, I. & De la Rosa-Rosero, F. (2020). Semi-automatic Construction of Video Game Design Prototypes with MaruGen. Revista Facultad de Ingeniería Universidad de Antioquia. (99), 9–20. [27] Charleer, S., Verbert, K., Gutiérrez, F., & Gerling, K. (2018). Towards an open standard for gameplay metrics CHI PLAY 2018 - Proceedings of the 2018 Annual Symposium on Computer-Human Interaction in Play Companion Extended Abstracts, New York, NY, USA. pp. 399–406. [28] Clegg, B. S., Rojas, J. M., &Fraser, G. (2017, May). Teaching software testing concepts using a mutation testing game. In 2017 IEEE/ACM 39th International Conference on Software Engineering: Software Engineering Education and Training Track (ICSESEET), Buenos Aires, Argentina. pp. 33–36. [29] Cooper, K. (2016). Proceedings of the 5th International Workshop on Games and Software Engineering, GAS, Austin, Texas, USA, May 16, 2016. ACM, ISBN 978-1-4503-4160-8 [30] Cooper, K., Scacchi, W., & Wang, A.-I. 2013. Proceedings of the 3rd International Workshop on Games and Software Engineering: Engineering Computer Games to Enable Positive, Progressive Change, GAS, May 18, 2013. IEEE Computer Society, San Francisco, CA, USA, ISBN 978-1-4673-6263-4 [31] Cooper, K. & Scacchi, W. (Eds.) 2015. Computer Games and Software Engineering, CRC Press, Taylor & Francis Group, Boca Raton, Florida. [32] Dalpiaz, F. & Cooper, K. (2020). Games for requirements engineers: Analysis and directions. In IEEE Software, vol. 37, no. 1, pp. 50–59, Jan.-Feb. 2020, doi: 10.1109/ MS.2018.227105450. [33] Daneva, M. (2017) Striving for balance: A look at gameplay requirements of massively multiplayer online role-playing games. The Journal of Systems and Software, 134. 54–75. [34] De Lope, R. P., & Medina-Medina, N. (2016). Using UML to model educational games. In 2016 8th International Conference on Games and Virtual Worlds for Serious Applications (VS-GAMES), Barcelona, Spain. pp. 1–4. [35] Debbiche, J., Lignell, O., Krüger, J., & Berger, T. (2019). Migrating Java-based apogames into a composition-based software product line. In Proceedings of the 23rd International Systems and Software Product Line Conference - Volume A (SPLC’19). Association for Computing Machinery, New York, NY, USA, 98–102. doi: 10.1145/3336294.3342361 [36] Delen, M., Dalpiaz, F., & Cooper, K. (2019, September). BakeRE: A Serious Educational Game on the Specification and Analysis of User Stories. In 2019 IEEE 27th International Requirements Engineering Conference (RE), Jeju, Korea (South). pp. 369–374.
Introduction to Software Engineering Perspectives
29
[37] do Prado, E. & Lucredio, D. (2015), A Flexible Model-Driven Game Development Approach, 2015 IX Brazilian Symposium on Components, Architectures and Reuse Software, Belo Horizonte, pp. 130–139, doi: 10.1109/SBCARS.2015.24. [38] Donovan, T. 2010. Replay: The History of Video Games, Lewes, Great Britain: Yellow Ant. [39] dos Santos, H. M., Durelli, V. H., Souza, M., Figueiredo, E., da Silva, L. T., & Durelli, R. S. (2019). CleanGame: Gamifying the identification of code smells. In Proceedings of the XXXIII Brazilian Symposium on Software Engineering. pp. 437–446. [40] Edwards, B. (2016), The forgotten world of BBS door games. PC Magazine, January 14, 2016, https://www.pcmag.com/news/the-forgotten-world-of-bbs-door-games [41] Emmerich, K. & Masuch, M. (2017). The impact of game patterns on player experience and social interaction in Co-located multiplayer games. In Proceedings of the Annual Symposium on Computer-Human Interaction in Play (CHI PLAY’17). Association for Computing Machinery, NewYork, NY, USA, pp. 411–422. doi: 10.1145/3116595.3116606 [42] Erdogmus, H., Medvidović, N., & Paulisch, F. (2018). 50 years of software engineering. In IEEE Software, vol. 35, no. 5, pp. 20–24, September/October 2018, doi: 10.1109/ MS.2018.3571240. [43] Fitzpatrick, A., Pullen, J., Raab, J., Grossman, L., Eadicicco, L., Peckham, M., &Vella, M. (2016), The best 50 video games of all time, Time Magazine, 2016, https://time. com/4458554/best-video-games-all-time/ [44] García, I., Pacheco, C., León, A., & Calvo-Manzano, J. A. (2020). A serious game for teaching the fundamentals of ISO/IEC/IEEE 29148 systems and software engineering– Lifecycle processes–Requirements engineering at undergraduate level. Computer Standards & Interfaces, 67, 103377. [45] Haendler, T., & Neumann, G. (2019). Serious refactoring games. In Proceedings of the 52nd Hawaii International Conference on System Sciences. [46] Hansen, D. 2016. Game On!: Video Game History from Pong and Pac-Man to Mario, Minecraft, and More, New York, New York: Macmillan Publishing Group. [47] Harris, B. 2014. Console Wars: Sega, Nintendo, and the Battle that Defined a Generation, New York, NY: HarperCollins Publishing. [48] Hernández Bécares, J. et al. (2017) An approach to automated videogame beta testing. Entertainment Computing, 18, 79–92. [49] Iftikhar, S. et al. (2015) ‘An automated model based testing approach for platform games.’ In 2015 ACM/IEEE 18th International Conference on Model Driven Engineering Languages and Systems (MODELS), Ottawa, Ontario, Canada. September 2015 IEEE. pp. 426–435. [50] LAN MAN Standards Committee of the IEEE Computer Society Approved: 26 June 1997 IEEE Standards Board. (1997) IEEE Standard for Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) specifications. IEEE Std 802.11-1997, 1– 445. doi: 10.1109/IEEESTD.1997.85951. [51] IEEE Standards for Local Area Networks: Carrier Sense Multiple Access With Collision Detection (CSMA/CD). (1985). Access method and physical layer specifications. In ANSI/IEEE Std 802.3-1985, doi: 10.1109/IEEESTD.1985.82837. [52] Internet Activities Board. (1988). RFC 1083 Official Protocol Standards. Available at: https://tools.ietf.org/html/rfc1083 [53] INCOSE Systems Engineering Handbook A Guide for System Life Cycle Processes and Activities (2015), 4th Edition, Hoboken, New Jersey: John Wiley & Sons.
30
Software Engineering Perspectives in Computer Game Development
[54] Ismail, N. & Lokman, A. (2020). Kansei engineering implementation in web-based systems: A review study. In Shoji, H. et al. (eds) Proceedings of the 8th International Conference on Kansei Engineering and Emotion Research. KEER 2020, Singapore: Springer. pp. 66–76. [55] ISO/IEC/IEEE International Standard. 2015. Systems and software engineering: System life cycle processes. In ISO/IEC/IEEE 15288 1st edition, 1–118, 15 May 2015, doi: 10.1109/IEEESTD.2015.7106435. [56] Junaidi, J., Anwar, N., Safrizal, W.H.L.H.S., & Hashimoto, K. (2018). Perfecting a video game with game metrics. Telkomnika (Telecommunication Computing Electronics and Control), 16(3), 1324–1331. [57] Kang, S. J., & Kim, S. K. (2015). Automated spatio-temporal analysis techniques for game environment. Multimedia Tools and Applications, 74(16), 6323–6329. [58] Karavolos, D., Liapis, A., & Yannakakis, G. (2017). Learning the patterns of balance in a multi-player shooter game. In Proceedings of the 12th International Conference on the Foundations of Digital Games (FDG’17). Association for Computing Machinery, New York, NY, USA, Article 70, pp. 1–10. DOI:10.1145/3102071.3110568 [59] Kasurinen, J., Palacin-Silva, M., & Vanhala, E. (2017). What concerns game developers? a study on game development processes, sustainability and metrics. In Proceedings of the 8th Workshop on Emerging Trends in Software Metrics (WETSoM’17). IEEE Press, pp. 15–21. [60] Kim, W. H. (2016) Efficient acceptance testing framework for interactive computer game applications. International Journal of Applied Engineering, 11(3), pp. 1815– 1819. [61] Kristiadi, D. P., Sudarto, F., Sugiarto, D., Sambera, R., Warnars, H. L. H. S., & Hashimoto, K. (2019). Game Development with Scrum methodology. In 2019 International Congress on Applied Information Technology (AIT). pp. 1–6. [62] Levey, S. 2010. Hackers: Heroes of the Computer Revolution - 25th Anniversary Edition, O’Reilly Media. [63] Liszio, S., & Masuch, M. (2016). Lost in Open Worlds: Design Patterns for Player Navigation in Virtual Reality Games. In Proceedings of the 13th International Conference on Advances in Computer Entertainment Technology (ACE’16). Association for Computing Machinery, New York, NY, USA, Article 7, pp. 1–7. DOI:10.1145/3001773.3001794. [64] Lui, R. W. C. , Geng, S., & Law, K. M. Y. (2017). Project management SPOC with animation. In Proceedings on 2017 IEEE 6th International Conference on Teaching, Assessment, and Learning for Engineering (TALE), Hong Kong, pp. 29–34, doi: 10.1109/TALE.2017.8252299. [65] Liu, S. et al. (2019) ‘Automatic generation of tower defense levels using PCG’. In Proceedings of the 14th International Conference on the Foundations of Digital Games. [Online]. 2019 ACM. pp. 1–9. [66] Loh, C. S., Li, I.-H., & Sheng, Y. (2016). Comparison of similarity measures to differentiate players’ actions and decision-making profiles in serious games analytics. Computers in Human Behavior, 64, 562–574. [67] Matallaoui, A., Herzig, P., & Zarnekow, R. (2015, January). Model-driven serious game development integration of the gamification modeling language gaml with unity. In 2015 48th Hawaii International Conference on System Sciences pp. 643–651. [68] Marin, C., Chover, M., & Sotoca, J. M. (2019). Prototyping a game engine architecture as a multi-agent system. In 2019 International Conference in Central Europe on Computer Graphics, Visualization and Computer Vision, pp. 27–34. [69] Maxim, B. R., Kaur, R., Apzynski, C., Edwards, D., & Evans, E. (2016, October). An agile software engineering process improvement game. In 2016 IEEE Frontiers in education Conference (FIE). pp. 1–4.
Introduction to Software Engineering Perspectives
31
[70] Meftah, C. et al. (2019) Mobile serious game design using user experience: Modeling of software product line variability. International Journal of Emerging Technologies in Learning. [Online] 14(23), pp. 55–66. [71] Milgram, P., Takemura, H., Utsumi, A., & Kishino, F. (1994). “Augmented Reality: A class of displays on the reality-virtuality continuum.”In Proceedings of Telemanipulator and Telepresence Technologies. pp. 2351–2354. [72] Miljanovic, M. A., & Bradbury, J. S. (2017 August). Robobug: A serious game for learning debugging techniques. In Proceedings of the 2017 ACM Conference on International Computing Education Research. pp. 93–100. [73] Mimura, Y., Tsuchiya, T., Moriyama, K., Murata, K., & Takasuka, S. (2020, July). UX design for mobile application of E-commerce site by using Kansei interface. In International Conference on Applied Human Factors and Ergonomics. pp. 641–647. [74] Mizutani, W. & Kon, F. (2020). ‘Unlimited rulebook: A reference architecture for economy mechanics in digital games.’ In 2020 IEEE International Conference on Software Architecture (ICSA). [Online]. March 2020 IEEE. pp. 58–68. [75] Molléri, J. S., Gonzalez-Huerta, J., & Henningsson, K. (2018). A legacy game for project management in software engineering courses. In Proceedings of the 3rd European Conference of Software Engineering Education. pp. 72–76. [76] Montenegro, C. H., Astudillo, H., &Álvarez, M. C. G. (2017, September). ATAM-RPG: A role-playing game to teach architecture trade-off analysis method (ATAM). In 2017 XLIII Latin American Computer Conference (CLEI). pp. 1–9. [77] Morisaki, S., Kasai, N., Kanamori, K., & Yamamoto, S. (2019) Detecting source code hotspot in games software using call flow analysis. In 20th IEEE/ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/ Distributed Computing (SNPD), Toyama, Japan, 2019, pp. 484–489, doi: 10.1109/ SNPD.2019.8935822. [78] Nagamachi, M. (Editor) 2017 Kansei/Affective Engineering, Boca Raton, Florida: CRC Press. [79] Nalepa, G. J., Gizycka, B., Kutt, K., & Argasinski, J. K. (2017). Affective Design Patterns in Computer Games: Scrollrunner Case Study. In FedCSIS (Communication Papers). pp. 345–352. [80] Naur, P. & Randell, B. Highlights (1968) Software Engineering: Report on A Conference Sponsored by the NATO Science Committee, Garmisch, Germany, 7th to 11th October 1968, NATOs http://homepages.cs.ncl.ac.uk/brian.randell/NATO/index.html. [81] O’Shea, Z. & Freeman, J. (2019). Game design frameworks: Where do we start? In Proceedings of the 14th International Conference on the Foundations of Digital Games (FDG ’19). Association for Computing Machinery, New York, NY, USA, Article 25, pp. 1–10. doi:10.1145/3337722.3337753 [82] Paduraru, C. and Paduraru, M. (2019). ‘Automatic difficulty management and testing in games using a framework based on behavior trees and genetic algorithms.’. In 24th International Conference on Engineering of Complex Computer Systems (ICECCS). pp. 170–179. [83] Paschali, E., Ampatzoglou, A., Escourrou, R., Chatzigeorgiou, A., & Stamelos, I. (2020). A metric suite for evaluating interactive scenarios in video games: An empirical validation. In Proceedings of the 35th Annual ACM Symposium on Applied Computing (SAC’20). Association for Computing Machinery, New York, NY, USA, pp. 1614–1623. DOI:10.1145/3341105.3373985 [84] Perez-Medina, J.-L. et al. (2019) ePHoRt: Towards a reference architecture for telerehabilitation systems. IEEE Access: Practical Innovations, Open Solutions. pp. 797159–797176.
32
Software Engineering Perspectives in Computer Game Development
[85] Peters, D., Calvo, R. A., & Ryan, R. M. (2018). Designing for motivation, engagement and wellbeing in digital experience. Frontiers in Psychology, 9, 797. [86] Politowski, C., Petrillo, F., & Guéhéneuc, Y. G. (2020, June). Improving engagement assessment in gameplay testing sessions using IoT sensors. In Proceedings of the IEEE/ ACM 42nd International Conference on Software Engineering Workshops. pp. 655–659. [87] Prasetya, W., et al. (2019, May). Having fun in learning formal specifications. In 2019 IEEE/ACM 41st International Conference on Software Engineering: Software Engineering Education and Training (ICSE-SEET). pp. 192–196. [88] Pyae, A., & Potter, L. E. (2016). A player engagement model for an augmented reality game: A case of Pokémon go. In Proceedings of the 28th Australian Conference on Computer-Human Interaction. pp. 11–15. [89] Quinones, J. & Fernandez-Leiva, A. (2020) XML-based video game description language. IEEE Access: Practical Innovations, Open Solutions. pp. 84679–84692. [90] Sandhu, A. & McCoy, J. (2019). A framework for integrating architectural design patterns into PCG. In Proceedings of the 14th International Conference on the Foundations of Digital Games (FDG’19). Association for Computing Machinery, New York, NY, USA, Article 49, pp. 1–5. DOI:10.1145/3337722.3341839 [91] Scacchi, W. & Whitehead, J. (2015) In Proceedings of the 4th IEEE/ACM International Workshop on Games and Software Engineering, GAS 2015, Florence, Italy, May 18, 2015. IEEE Computer Society 2015, ISBN 978-1-4673-7046-2 [92] Scacchi, W. & Cooper, K. (2015). Research Challenges at the Intersection of Computer Games and Software Engineering. In Proceedings of the 10th International Conference on the Foundations of Digital Games, FDG 2015, Pacific Grove, CA, USA, June 22–25, 2015. https://dblp.org/rec/bib/conf/fdg/ScacchiC15 [93] Schatten, M. et al. (2017). Towards an agent-based automated testing environment for massively multi-player role playing games. In 2017 40th International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO). pp. 1149–1154. [94] Time Magazine cover article (1982). http://content.time.com/time/covers/0,16641, 19820118,00.html January 18, 1982, Vol. 119, No. 3. [95] Shi, W., Kaneko, K., Ma, C., & Okada, Y. (2019). A framework for automatically generating quiz-type serious games based on linked data. International Journal of Information and Education Technology, 9(4), pp. 250–256. [96] Sierra, M., Pabón, M., Rincón, L., Navarro-Newball, A., & Linares, D. (2019). A Comparative Analysis of Game Engines to Develop Core Assets for a Software Product Line of Mini-Games. In International Conference on Software and Systems Reuse. pp. 64–74. [97] Söbke, H. & Streicher, A. (2016) Serious games architectures and engines. In Dörner R., Göbel S., Kickmeier-Rust M., Masuch M., Zweig K. (eds.) Entertainment Computing and Serious Games. Lecture Notes in Computer Science, vol 9970. Cham: Springer. doi: 10.1007/978-3-319-46152-6_7 [98] Stahlke, S. et al. (2019) ‘Artificial playfulness’. In Extended Abstracts of the 2019 CHI Conference on Human Factors in Computing Systems. pp. 6984–6986. [99] Su, Y. Game Analytics Research: Status and Trends (2020) Lecture Notes on Data Engineering and Communications Technologies, vol. 41, pp. 572–589. [100] Summerville, A., Snodgrass, S., Guzdial, M., Holmgård, C., Hoover, A. K., Isaksen, A., & Togelius, J. (2018). Procedural content generation via machine learning (PCGML). IEEE Transactions on Games, 10(3), pp. 257–270.
Introduction to Software Engineering Perspectives
33
[101] Taborda, J., Arango-López, J., Collazos, C., Vela, F., & Moreira, F. (2019) Effectiveness and fun metrics in a pervasive game experience: A systematic literature review. Advances in Intelligent Systems and Computing, 932, pp. 184–194. [102] Vakaliuk, T., Kontsedailo, V. V., Antoniuk, D. S., Korotun, O. V., Mintii, I. S., & Pikilnyak, A. V. (2019). Using game simulator Software Inc in the Software Engineering education. In Proceedings of the 2nd International Workshop on Augmented Reality in Education, Kryvyi Rih, Ukraine, March 22, 2019, No. 2547, pp. 66–80. [103] Valente, L. et al. (2017) Mapping quality requirements for pervasive mobile games. Requirements Engineering, 22 (1), pp. 137–165. [104] Varvaressos, S. et al. (2017) Automated bug finding in video games. Computers in Entertainment: CIE, 15 (1), pp. 1–28. [105] Vegt, W., Bahreini, K., Nyamsuren, E., & Westera, W. (2019). Toward reusable game technologies: Assessing the usability of the RAGE component-based architecture framework. EAI Endorsed Transactions on Serious Games, 5(17). doi: 10.4108/ eai.11-7-2019.159527 [106] Walk, W., Görlich, D., & Barrett, M. (2017). Design, dynamics, experience (DDE): An advancement of the MDA framework for game design. In Game Dynamics. pp. 27–45. [107] Wang, Y., Ijaz, K., Yuan, D., & Calvo, R. (2020). VR-Rides: An object-oriented application framework for immersive virtual reality exergames, Software: Practice and Experience, 50 (7), pp. 1305–1324. [108] WAP Forum. June 1999. Wireless Application Protocol, WAP White Paper, Mountain View, California. [109] Whitehead, J. & Lewis, C. (2011) Workshop on games and software engineering (GAS 2011). In Proceedings of the 33rd International Conference on Software Engineering (ICSE '11). Association for Computing Machinery, New York, NY, USA, 1194–1195. doi:10.1145/1985793.1986042 [110] Wolf, M. 2007. The Video Game Explosion. Greenwood, Westport, CT. [111] Zahari, A. S., Ab Rahim, L., Nurhadi, N. A., & Aslam, M. A. (2020) Domain-specific modelling language for adventure educational games and flow theory. International Journal onn Advanced Science Engineering Information Technology, 10(3), pp. 999–1007. [112] Zheng, Y. et al. (2019) ‘Wuji: Automatic online combat game testing using evolutionary deep reinforcement learning.’ In 2019 34th IEEE/ACM International Conference on Automated Software Engineering (ASE). [Online]. November 2019 IEEE. pp. 772–784. [113] Zhu, M., Wang, A. I., & Trætteberg, H. (2016, November). Engine-Cooperative Game Modeling (ECGM) Bridge Model-Driven Game Development and Game Engine Toolchains. In Proceedings of the 13th International Conference on Advances in Computer Entertainment Technology. pp. 1–10.
Section I Emerging Research on Serious Games for Software Engineering Education
2
Case Studies and Practices in Local Game Jam Software Development Organization A Software Engineering Perspective Walt Scacchi University of California, Irvine
CONTENTS 2.1 Introduction����������������������������������������������������������������������������������������������������� 38 2.2 Organizational Forms for Local Game Jams and Related Game Creation Events����������������������������������������������������������������������� 40 2.2.1 Other Variations����������������������������������������������������������������������������������� 42 2.3 Local Game Jam and Game Production Field Studies������������������������������������ 42 2.3.1 Intramural Game Jams at UCI VDGC������������������������������������������������ 43 2.3.2 VGDC Intramural Serious Game Jams����������������������������������������������� 45 2.3.3 Inter-Mural Game Development Showcase���������������������������������������� 46 2.3.4 Capstone Game Development Project Course������������������������������������ 49 2.3.5 International Game Development Workshop for Global Game Software Engineering Education���������������������������� 50 2.4 Observations, Lessons Learned and Learnable Lessons for Game Jam Organizational Forms�������������������������������������������������������������� 51 2.5 Conclusions����������������������������������������������������������������������������������������������������� 54 Acknowledgments���������������������������������������������������������������������������������������������������� 55 References���������������������������������������������������������������������������������������������������������������� 55
37
38
Software Engineering Perspectives in Computer Game Development
2.1 INTRODUCTION Many students and independent game developers participate in computer game development competitions, hackathons [41], or game jams [12, 17, 22, 23, 29, 34, 42]. These team-based game making efforts typically focus on clean-sheet production of a playable game usually in a limited time frame, like 24–96 hours though shorter and longer competitions have been engaged. Sometimes these jams have external for-profit or non-profit sponsors, who in turn may offer financial or technology product rewards to motivate participants to excel. Other times, jams offer no tangible rewards, but instead focus on going “for the win,” résumé building, demonstrating game development competency, and earning local geek status and shared learning experience as the desired outcome. In any case, game jams offer the potential to serve as innovation factories that can rapidly prototype many games in a short period of time [17, 21]. Game jams vary in geographic scope, from global game jams to regional or venue-specific (hereafter, local) game making events. This paper focuses attention on local game jams and game creation efforts, with an eye on examining organizational practices and possibilities in such game development activities. Local game jams, game development competitions, and game making events are those where game developers commonly know one another across teams, and celebrate the sturm und drang of their mutual game development successes and problems together. As such, such locality more readily embraces convivial as well as competitive game development. An interesting set of research potentials arise associated with game jams, perhaps most relevant to empirical studies of alternative computer game software engineering (CGSE) processes, practices, methods, or tools used. For example, within local intramural game jams, it may be possible to structure and balance the game development teams by team size, game developer roles, and SE skill level from students at hand. Students can indicate their skill level and developer role preferences, then have participants randomly assigned to teams in ways that balance team size, role, and skill level. This can mitigate against pre-formed teams with established collaborators, high skill distribution, and relatively mature game development capabilities. How might such teamwork structures affect how games are made, or the quality of the game products that are produced? Such questions require empirical study, and game jams may provide the venues for such studies to be conducted. Short-duration jams mitigate against the consequences of team failure or participant drop out, and instead make these events more of a CGSE learning experience. In this way, in addition to focusing on game production, the overall game jam can serve as a “field site” where selected CG design, SE processes, and technologies [e.g., 16, 17, 20, 29, 30, 31] can be comparatively investigated, following empirical SE approaches introduced more than 25 years ago [2, 3]. Such field sites can allow for informal or systematic empirical study of teams using a new game software development kit (SDK) for indie game development [9, 30], or development process/ technique (e.g., SCRUM, agile development [42], or game modding [27, 28]) versus those who do not; or those who produce traditional SE documents (requirements
Local Game Jam Software Development Organization
39
specifications, architectural designs, test plans) and follow SE processes for their game [42] versus those who just focus on game design [32] or “minimum viable product” (MVP) methods. Intramural game jams so structured may therefore be well-suited for longer durations (e.g., from days to weeks), though ultimately this is a game jam design choice. Game making competitions may stress short duration and co-location, along with targeted game production on a topic that is announced at the beginning of the competition. Inter-mural game jams, those open to teams from different schools, may not be so readily structured or balanced at little cost, but instead may address other CGSE questions that better match their natural field organization, geographic distribution, and project heterogeneity. However, there is no inherent requirement that game development competitions must be of short duration, as it is possible to create or find examples of those extending for years. More generally, game jams offer the opportunity to organize, design, and conduct empirical studies in CGSE that can inform both new game design practices or processes, as well as new SE practices and technologies [7, 42]. Some game SDK vendors have sponsored game jams that focus on participating team’s usage of specific proprietary products or platforms (e.g., Epic Games’ Make Something Unreal and Microsoft’s Imagine Cup: Game Design). These competitions can be used to address CGSE research questions in ways underutilized in SE research, and thus highlight similarities and differences between traditional approaches to computer game design [32] and software engineering [16, 31]. Ultimately, this can mean that game jamfocused SE can be viewed as a competitive team-based sport activity that can be fun for students, as well as structured to support careful empirical study [14, 29, 31], rather than SE being a business endeavor to produce application systems hosted on back-end infrastructures accompanied by voluminous documents that few will ever read. Finally, it is also possible to recognize new, under-explored ways and means for organizing game jams for further empirical study. For example, local computer game jams may be organized and designed as a kind of meta-game—a game about (making) games. Such game making events can be organized as a game, whose form resembles a game-based course, where student participants earn points and rewards while leveling up to a higher level of academic proficiency [33]. Such jams may take organizational forms whose goal is to structure the outcomes (i.e., the games produced) to embody certain functional features or the CGSE production processes to reward accumulative levels of progress achieved or skills mastered by different teams (“leveling up”), rather than just leading to winning and losing teams. Alternatively, game jams may be both experimentally modeled and studied as teamwork structured, CGSE role-playing games [18]. Last, game jams may focus on exploring cross- cultural or global CGSE development processes, whereby teams involve participants who collaborate across time, space, or cultural distance, via global CGSE processes and practices [1]. With these alternative forms for organizing game making events, it is now appropriate to identify and briefly describe different kinds of local game jams that we have been able to put into practice and study.
40
Software Engineering Perspectives in Computer Game Development
2.2 ORGANIZATIONAL FORMS FOR LOCAL GAME JAMS AND RELATED GAME CREATION EVENTS Local game jams may be located in academic settings of different kinds. This may be especially important to academic scholars in game design or game studies, who seek innovative ways and means for conducting some form of empirical or experimental approaches to game development. Much like team sports in schools, game development competitions can be organized as intramural (within school) and inter-mural (across schools). These team-based game development efforts can be undertaken in ways that complement formal CGSE (or just SE) educational principles and practices [5, 6, 29, 34, 40]. Sub-types of intramural and inter-mural game jams can be identified, for example, when external sponsors are involved and tangible rewards are offered as incentives to motivate game jam participants, in contrast to jams where there are no external sponsors or tangible rewards so that game developers focus on symbolic rewards and convivial learning experiences as their motivation to participate. Such variations allow for examination of whether and how external versus internal motivations affect the resulting games, local game development practices, teamwork, and the satisfaction of overall participant experience [22, 23]. Capstone project courses are growing in popularity in computer game design, software engineering, or computer science degree programs. These projects are organized and managed by faculty who determine constraints on matters like team composition, choice of game SDK, or game deployment platform (e.g., games for Android smartphones, web browsers, or PCs). Alternatively, faculty may allow their students to make some/all of these choices. The faculty often serve as coaches that mediate and motivate student teams to be creative and collaborative, yet assure their need to develop and deliver a complete project, along with required documentation, demonstration, and final presentation. Sometimes these project courses benefit from external or non-academic project sponsors who see capstone project courses as a low-cost means for prototyping new game concepts, utilizing emerging technologies, or making serious games targeted to some appropriate application problem domain. Another variation of the capstone project can be seen in tracing back the roots of jams to earlier exemplars in (non-game) domains like “Anijams” introduced and popularized for years within the film animation production community (not to be confused with Ani-Jams which are more recent community events for anime fans). Legendary film animator Marv Newland and his International Rocketship Ltd. production company created a collaborative animated film making project where 22 animators in different locations each created a short-animated film segment, given only the first and last key frame (graphic image) that they would then create the inbetween frame sequences that would be included in the final composite film [19]. This collaborative development project effectively employed a common baseline visual outline that served as an animation storytelling architecture via the sequence of key frames. The project was envisioned to accommodate the independent, parallel creation of in-between animation sequences. A game-oriented repurposing of this approach might take the form of a multi-team game development project where the game segments conform to a common software architecture design and run-time
Local Game Jam Software Development Organization
41
platform environment [39]. These semi-autonomous games could then be treated as modular architectural plug-ins, can interoperate to exchange gameplay data/content assets across game modules. Game development workshops are another kind of game making event that can explore or structure the development of game products and artifacts. Workshops can also experiment with and determine the practical efficacy of different tool/SDK selections, as well as the enactment of CGSE processes, project team forms, and work practices. Again, these workshops may be intramural or inter-mural, though it is easier to see them as open in some way to participants who are not specifically seeking academic coursework credit, but more of an advanced or focused game development experience. Furthermore, such workshops may be externally sponsored and may invite international participants, perhaps in ways that resemble the oncepopular NATO Summer School Workshops that attract graduate or post-doctoral students in local, scientific, or policy problem-solving programs. Intensive workshops rely on the organizers to provide intellectual leadership and technical guidance for the participants. In practical terms, a game development workshop may differ from an intramural game jam or capstone project through administrative and management choices. For example, a game development workshop may involve the engagement and direction of multiple project leaders or subject-matter experts who serve to formulate, plan, and guide those activities engaged by workshop participants. Workshops may have an explicit curriculum and process plans that articulate a schedule of activities that correspond to the delivery and engagement of participants with specific topics. A CGSE workshop may therefore have faculty or industry experts who lecture (perhaps using a “flipped class” format) on a topic like a selected game software architecture and run-time environment that is conceived to demonstrate and embody SE concepts like mini-games as architectural plug-ins; extension mechanisms that support multiple target platforms and modding via the use of multiple databases; and use of specified interfaces (APIs) to online banking/ commerce services for encrypted user-specific transactions. Addressing such issues is commonly beyond what can be done in short-duration game jams, or game making events where participant teams are open and ad hoc, such that game makers determine the design of everything in the game. Game development showcase festivals may be organized as either intramural, inter-mural, or open to independent game developers (whether or not affiliated with a university, school, or game studio). Game showcase festivals are organized in ways similar to film festivals (e.g., Sundance, Tribeca, and Berlin), where game development processes, tools used, team structures, artifacts produced, budget, schedules, etc., are all non-issues in the festival, which instead focuses near-exclusive attention to the games as products (or online services). With a product-centered focus, game developers focus attention to winning awards or recognition, as well as possible publication, distribution, or investment deals from third parties (now including crowdsourced funders). So game showcase festivals can marginalize the significance of game development practices, other than to telegraph that winning teams must somehow be creative and technically competent in game development, but with little/no knowledge of anything learned or experienced by the game developers or of game development.
42
Software Engineering Perspectives in Computer Game Development
2.2.1 Other Variations Beyond the game jam types just identified, it is also possible to classify other recurring models of multi-game production competitions as game jam variations. For instance, game modding communities and portals routinely host contributed games/ mods developed by teams, which are then played, reviewed, and ranked by other online game players. Examples of mod development competitions include the Thief Modding Contest, Make Arma Not War, and Star Citizen Modding Competition. Next, there are multiple, independent yet coordinated team research projects funded by external sponsors or government agencies, like the DARPA Crowdsourced Formal Verification initiative called Verigames [38] in which a number of research-grade university teams undertook sustained, multi-year efforts to produce, deploy, and assess games that embody different approaches to crowdsourced gameplay for verifying large software systems (e.g., the Linux Kernel). Last, there are also competitions that focus on game materiality, such as costumes and wearables as game controllers [35], and case modding. Examples of the latter include Nvidia’s Mod24, a full day long case modding competition, and Cooler Master case modding competition. So there is no shortage of formats, platforms, and venues that address different configurations of game development competitions. Any or all of the above categories can be classified as collective, participatory teams working in parallel to make games or game technologies. Most often, attention is directed to the products of game development efforts, principally the games produced. However, game jams can also focus attention to the comparative study of game development practices, collaborative game software development teamwork processes, evaluation of game development artifacts (game design versus run-time implementation versus game post-mortem), efficacy of game development tools/ SDKs employed, or some combination of these socio-technical elements. Game competitions can also be extended to support other CGSE challenges like team-based game play-testing jams, or be aligned with game playing competitions, depending of the participants sought, and the audiences (or external sponsors) to be embraced. With these different organizational forms for game making events, we now turn to describe a set of five field studies where we have observed multi-team game making events.
2.3 LOCAL GAME JAM AND GAME PRODUCTION FIELD STUDIES Game development competitions can arise in diverse settings with different constraints and game development affordances. Five different kinds of game making efforts have been systematically observed as field studies in multi-team game development. These include (a) local intramural game jams hosted by a student-run game development club, just for the fun of the experience; (b) local intramural serious game jams with external sponsor and post-jam development contract to the overall winning team; (c) a regional inter-mural game showcase with teams from different schools, along with multi-school teams; (d) intramural game capstone project course for academic credit; and (e) an international multi-team game development workshop focused on producing game modules for incorporation within a common game
Local Game Jam Software Development Organization
43
software architecture. Each is described in turn, then comparatively analyzed to identify observations, lessons, and opportunities for how to organize and design game jams. Such results may facilitate playful game making learning experiences and potentially innovative games, as well as do so in ways that generate new, empirically substantiated game software engineering research findings.
2.3.1 Intramural Game Jams at UCI VDGC Video game developer clubs (VGDCs) are up and running at many colleges and universities. As a student-run venture, they can elect to host game jams as extracurricular activities that are non-academic (no faculty mentors, no course credit) and open to all students who voluntarily attend. The UCI VGDC has students from many academic majors, including those not directly connected to computer games (e.g., biological sciences), as well as students from off-campus groups (e.g., nearby high schools and technical schools). Since 2009, the VGDC has organized and run game jams, commonly three times per year. As the VGDC annual membership fluctuates between 60–100+ students per year, the game jams also serve as a recurring, core social event for the club that focuses on student teams building games on short schedules (weekend or week long) on a single topic or theme (e.g., electricity, health, and friction) that is randomly selected at the beginning of the game jam. See Figure 2.1 for an example. Dozens of games have been developed and demonstrated across the jams, with 5–12 games presented at the end of each jam. The resulting games, as suggested by Figure 2.2, are demonstrated live before a public audience and panel of faculty judges. The live demo presentation includes a brief post-mortem, that highlights team-specific lessons learned. As Grossman [10] reports, compiled post-mortems reiterate common problems with team projects, such as running out of time, need to continually reduce
FIGURE 2.1 Online photo documentation from intramural game jam held at UCI in 2011.
44
Software Engineering Perspectives in Computer Game Development
FIGURE 2.2 Online video documentation that demonstrates the operation of a game produced in an intramural game jam at UCI.
game design scope [cf. 32], and a small number of team members underperforming or failing to perform as promised. The volunteer faculty judges evaluate the games on subjective criteria jointly identified with the VGDC (e.g., game graphic design, interesting use of sound/music, user experience, and efficacy of the post-mortem confessional). The judges are asked to provide overall rankings, as well as brief written feedback for the teams. The winning team receives hardy applause from the audience, and pats on the back, along with symbolic pride points and improved local reputation. The success and frequent recurrence of these local game jams as a fun but challenging way to rapidly prototype new games even caught the attention of local news media [11]. Over time, and in concert with student-voiced demands, the jam organization has evolved to forms where team members are chosen at random by role preference (programmers, artists, musicians, modelers, producers, writers) and balanced. This means teams have roughly the same number of team members and skill sets and do not necessarily have prior game development teamwork relationships. This helps to ameliorate that chance for established collaborators teaming up time and again as an uneven competitive advantage, since VGDC students prefer “friendly competitions” that stress challenges that are bounded and learning-oriented, rather than cutthroat, winner-take-all competitions. This is a recognition that most competitions produce few/one winner, but mostly produce “losers,” which goes against the spirit of a learning environment, where gentle (and sometimes frequent) failure leads to improvement. Student teams are free to choose the game development tools and techniques they want to use. So, teams may choose complex game SDKs like Unity or Unreal Development Kit, or something very generic like open source Java game development libraries or now deprecated Microsoft XNA libraries (which seem to reinforce development of 2D platformer games), as well as popular cloud-based software
Local Game Jam Software Development Organization
45
tools/services like Slack, Discord, GitHub, Steam, and others. In general, students have chosen not to utilize other popular game SDKs [30] like Aurora for Never Winter Nights, GameMaker, Construct 2, extensible virtual worlds like Second Life, or open-source game software approaches [25, 28], out of a lack of prior experience, lack of interest, or belief that these are not “professional” game development environments. Tool choice is generally decided by team programmers based on their prior experience, current preferences, or trust in declarations by reputable game developers communicated via online social media. Teams also decide which game/software development artifacts to produce. Most common are game design documents, but also shared are persistent chat transcripts, online/in-game user tutorials, and game jam demo presentation slide decks. Finally, some teams find sufficient self-interest and enthusiastic play-testing responses from users that they elect to continue to develop the game after the jam, with the goal of publishing the result in an online game store like Microsoft Xbox Live. In contrast, student teams do not produce CGSE documents like explicit functional or non-functional requirements, nor systematic test and integration plan specifications [4]. Thus, it is unclear if this is a missed opportunity or just something the students find irrelevant to their game jam development efforts [31]. Last, VGDC students have stated their interest to prefer to participate in these local intramural game jams where they can know the other students developing games and participate in a local game making scene. This is in contrast to their seeking the same level of participation in events like the Global Game Jam. In previous years, VGDC students indicated they felt the GGJ was too remote and weakly engaged, and open to teams with independent game development “ringers” who may unbalance the game jam playing field competition, thus focused more attention to prize winning (and thus to producing mostly losing teams and games that are lost in large under-differentiated game submission repositories). Similarly, the UCI VGDC students have multiple game jams to elect to participate in, including those described below. Nonetheless, in 2015, the VGDC mobilized participation and game submissions for the GGJ, due in large part to the leadership and encouragement of a new UCI faculty member focused on computer games. More recent VGDC efforts at UCI are found on their online sites hosted by Google Sites, Facebook, LinkedIn, GitHub, and Itch.io, along with their Twitter and Discord channels. Beyond this, may other collegiate clubs that participate in local or regional game jams can be found via the Student Game Developer Alliance website.
2.3.2 VGDC Intramural Serious Game Jams As the VGDC developed a reputation for engaging undergraduate students in independent game development outside of their coursework, outside industry, academic, and government partners have sought to sponsor game jams aligned with their institutional interests. In this regard, these outside groups want to sponsor a jam that produces serious games that address their interests. Often at times, these jams are envisioned to allow the external sponsor to reach out to students as way to encourage student innovators/entrepreneurs or as a form of job recruitment. These arrangements may bypass the engagement of the host school, their faculty, external grants
46
Software Engineering Perspectives in Computer Game Development
FIGURE 2.3 Screenshots from the HeartRace game initially produced for a UCI serious game jam to encourage healthy heart lifestyle choices. Support for this game’s ongoing development was provided by the American Heart Association, Orange County Chapter.
development offices, and student placement services, but in general no major administrative problems or resistant academic politics seem to surface. As an example, the local chapter of the American Heart Association approached the VGDC to sponsor a game jam whose game efforts would focus on the topic of “healthy hearts” and related healthy lifestyle choices. Six teams completed and presented their games after a week-long jam for review and evaluation by panel of judges from the AHA, local game industry, and UCI faculty. The winning team then received a cash prize, along with a 6-month contract to further develop the game for possible deployment and release under AHA sponsorship. See Figure 2.3 for example screenshots from this game.
2.3.3 Inter-Mural Game Development Showcase Based on part from the successes and experiences of the intramural game jams at UCI, and the growing participation of students from other nearby trade schools, colleges, and universities, an effort was mounted to elevate these regional game jams into an inter-mural form. As a number of faculty, both local to UCI and nearby, were also active in the local chapter of the IEEE Computer Society (as well as the Independent Game Developers Association) that hosted a special interest group in Games (SIGG), and the local IEEE advocates were searching for new ways to enlist students into their profession (mostly populated by mature engineers), then a relationship with SIGG
Local Game Jam Software Development Organization
47
FIGURE 2.4 Online website banner for the 2013 edition of IEEE GameSIG Intercollegiate Game Showcase held at UCI.
emerged to establish and host such an inter-mural event. See Figure 2.4 for example. However, as different schools have different levels of student interest and coursework in computer games (e.g., UCI has 4-year CS degree program with a dozen game focused courses, while other schools may offer only 1–2 courses in game design), the SIGG people decided to organize the game jam in a manner more like a game showcase (similar to IndieCade or film festivals) where development time or team size is not constrained, rather than as is common in a time-limited game jam. This means student teams would be unbalanced, could take as long as they wanted to make their game, engage whatever tools and techniques they found appropriate, and even be able to submit games that were projects in a capstone project course, or those submitted from other game jams unrelated to this competition and its sponsors. The showcase requires that game software and content assets must also be packaged and posted on servers, for download and installation. Also, teams had to submit a 5-minute or less video of the game team and design pitch (present the game concept) and recordings of live play sessions, as seen in Figure 2.5. Numerous example game demo videos can be found on YouTube via a search for “IEEE gamesig” and “IEEE game sig.” Game teams compete for recognition from game industry veterans, who select the best games and game demo/presentations, to determine the winners in different categories (e.g., best mobile game and best game overall), as suggested in Figure 2.6. The industry judges include executives and lead designers from large multimillion-dollar game studios, long-term independent game developers, or leading game artists/musicians, all of whom are familiar and experienced with publishing successful, moneymaking games. The judges’ final decisions then determine prize winners. Noteworthy here is that some of the overall game showcase winning game
48
Software Engineering Perspectives in Computer Game Development
FIGURE 2.5 Sample of game demo video deliverables associated with the 2018 edition of the IEEE GameSIG Intercollegiate Game Showcase.
FIGURE 2.6 Photo documentation from a 2015 inter-mural game showcase highlighting live game demos (upper left), a winning team (upper right), game project presentation, and local game industry participants from Blizzard Entertainment (lower left and right).
Local Game Jam Software Development Organization
49
teams have gone on to receive external investment from either angel investors who attend the showcase and observe audience reactions/interest, or from crowdfunding sites (Kickstarter, Indiegogo, Patreon, and others). Such investments thus help to evolve and transform a student team game jam result or game project into a new venture. Again, the boundaries and outcomes of such inter-mural games showcase competitions merit further consideration and study. As this form of game development competition is an inter-mural activity that engages students teams from multiple university or colleges, then it may be natural to ask if the competition can be formalized into something approaching an intercollegiate sport—in effect, treating game development as a team sport [29]. While scholars like Taylor [36] and others have examined the socio-technical, material, and creative dynamics of esport events that focus on competitive game play, it is probably too early to expect that inter-mural game jams may get elevated to full-fledged collegiate sport. But maybe our expectations are too low or misdirected. Alternatively, inter-mural game jams can grow to incorporate teams that span multiple schools that are geographically dispersed, thus affording participants firsthand experience in distributed, multi-site project work that relies on shared online information, social media, and related information repositories, all of which are key to the future of SE work in the Internet age [1].
2.3.4 Capstone Game Development Project Course Faculty and students are increasingly familiar with capstone project courses whose goal is to unify and demonstrate what students have learned through their prior coursework. Such project courses can span common academic periods like 10-week quarter or 15-week semester, or longer (UCI’s game capstone project courses for majors are now for 20 weeks). These project courses are not explicitly organized as formal competitions with winners and losers, but more like long-duration intramural game jams, where sustained focus, hard work, collaborative development, and selfsatisfaction are all desirable elements of open, multi-team software development project work [26, 27]. Capstone projects have been employed in SE education (SEE) programs for decades. SE project courses first appeared in the 1970s (cf. “Programming in the Large” c. 1975) and began to flourish in the 1980s, along with the establishment of the Software Engineering Institute at Carnegie-Mellon University in the mid-1980s. Part of the early charter of the SEI was to encourage and advance SEE as a national priority. Different project organizations and SE processes were thus encouraged and celebrated in research papers addressing SEE. Some educators further specialized in operationalizing their SE project courses as research test beds where experiments or other empirical studies of SE could be performed, analyzed, compared, and published [2, 3]. Among the SE practices that were subject to empirical study were the use of rapid prototyping versus conventional approaches to software development [3, 21], and utilization of balanced student teams developing common formal and informal software development artifacts, with/without reusable exemplars, on a short schedule [2]. Other SEE projects employed either a common software architecture
50
Software Engineering Perspectives in Computer Game Development
[39], or one with plug-in modules for each team to develop [40], or else a common software development infrastructure of tools/services [24]. As game capstone project courses are still emerging, as is the community of practice focusing on CGSE, then it is still common to find such courses organized as independent student teams each pursuing their personal interests in their game development efforts. Again, this also reflects popular practices that are used in independent game festivals and inter-mural game showcases. At UCI, the first capstone project course featured three student teams, while the most recent effort featured nine teams, reflecting the growth of this local program.
2.3.5 International Game Development Workshop for Global Game Software Engineering Education In Fall 2014, the author and other faculty at UCI were approached by a group of companies and universities from South Korea and China. They were interested in us developing and offering a 6-to-8-week long computer game development workshop for international students from these nations. The resulting Workshop was planned for 2016, so the description that follows is for the effort as designed, rather than an historical record describing what it was. But like some software development projects, and many game development projects, the project was cancelled early on due to schedule volatility, recurring delays, and multiple budgetary reductions [10]. Thus, this is a case for the speculative design [8] of an international game development workshop, rather than of accomplished design outcome for such. The Workshop was conceived and organized as a multi-week summer school, in the spirit of the NATO Summer Schools organized about new scientific research topics. The 30 or so participating students are grouped in near-balanced teams (to accommodate the uncertainties in student experience and game development capabilities), with six or so teams anticipated. There are five academic faculty who are co-organizing and co-teaching the Workshop, which is to be run as a large game development studio with all teams working on a common game architecture and target deployment platforms (e.g., online Web and mobile devices). The faculty have also recruited more than a dozen game industry veterans to serve as guest lecturers on specific topics, or who may participate in student project reviews. A common software architecture has been designed, along with a common game development infrastructure of tools to be used, and game development artifacts to be produced for faculty and peer review. The game software architecture is structured around plug-in game modules, where each module provides, at minimum, a complete mini-game experience, as well as exchange and interoperation of game play assets (in-game resources and play scores that can move across modules, accumulate, and persist). The game development infrastructure assumes multiple loosely coupled repositories for sharing game software files (versioned, as done using GitHub), in-game content/ assets, models, and textures (versioned); user play data (in-game character customizations, resource holdings, and scores); and anticipated online store repositories for game distribution. So, what kind of a game making event is this Workshop? Is it a game jam? First and overall, the Workshop is inspired by the alternative game jam organizational
Local Game Jam Software Development Organization
51
forms described above, with elements drawn from intramural jams (all local teams, working collocated and in parallel on same schedule, producing common deliverable types), inter-mural jams (students are from different schools in their home countries, though student teams are expected to be pre-formed at the home institution before the Workshop), and capstone project following the SE project forms that utilize balanced teams working on recurring short (bursty) schedules to produce targeted game artifacts (design documents, character designs, etc.). In addition, this Workshop enables one form of global software engineering education (SEE), specifically global game software engineering, where cross-cultural game developers and project managers must rapidly learn how to productively work together [1], rather than not engaging in cross-cultural game development challenges, as is often the practice in global game jams. Second, competition in the form of comparative assessment of each participating team’s effort, along with team pride, can be utilized as a constructive motivation [12, 22, 23]. Also, as the teams are expected to produce games (or game modules) that will be part of a commercial release, then the short product development cycle tends more toward a game jam, where severe time limitations help to encourage (or force) timely decision-making on creative or technical choices. Last, whether this style of game jam is the one that relies on practices such as gamified coursework [33]; advanced CGSE education [5, 6, 15, 16, 40]; rapid prototyping [3, 17, 21]; or software development work forms that resemble industrial “playbour” [13, 38] rather than hard work and fun that is mutually beneficial and exploitative for both game studios/sponsors and game modders/hobbyists [27]; etc. is an open question for further study. Suffice to say that other large software development projects have followed such development organizational forms, that the effort merits investigation as yet another viable method for a game jam, in this case, one that focuses all teams to contribute to the development of a single overall game experience, but one that is factored into plug-in functional modules that can share and exchange common game play resources and assets.
2.4 OBSERVATIONS, LESSONS LEARNED AND LEARNABLE LESSONS FOR GAME JAM ORGANIZATIONAL FORMS Based on a comparative review of the five field studies presented above, a number of observations follow from these five field studies of local game jams. First, local game jams are a promising venue for empirical studies of game development and software engineering. Many different kinds of configurations for organizing such jams have been identified. More potential configurations and reconfigurations therefore seem likely. Similarly, there is more potential opportunity to look for ways and means for structuring such software development competitions, or for analytically framing such competitions as a team sport [29], and thus also suitable for further gamification. Second, most game jams, whether local or global, have primary emphasis on the product of the development effort—the game produced—while discounting or ignoring attending to the affordances and capabilities that are rendered in game development artifacts (e.g., game design documents; persistent online game developer chat
52
Software Engineering Perspectives in Computer Game Development
transcripts to codify the knowledge and team logic underlying a game’s development; choices made regarding use of game development tools or SDKs as a mediating factor in game production; and whether development teams are balanced by role and team size versus ad hoc and open). Game development competitions also represent a relatively unexplored domain for empirical studies of collaborative software development teamwork [15, 25], particularly those that rely on online artifacts (e.g., game design documents, persistent chat transcripts, game screen layout and artwork mockups, and game mods) moving within/across shared repositories and social media [26], which may therefore represent a promising approach to renovate traditional non-game software development projects. Third, local game jams accommodate organizational design variables that may be more readily structured and (experimentally) controlled compared to open global game jams. This is not to say that local is better than global, but instead to draw attention to what variables can be addressed in different or comparative studies of game development competitions. Global game jams are well-suited to studies of crosscultural game development practices, and perhaps readily extended to also investigate global SE or global CGSE issues. Much remains to be investigated, observed, analyzed, and compared here. So spatial, temporal, and cultural distances within or across GGJ teams, or within international game development workshops, seems like another opportunity to explore or put into practice. Fourth, game development competitions are yet (?) to embrace participation in large-scale game development (e.g., MMOG) whereby teams compete based on their contribution to an overall established game software and asset/content architecture as the common focus for all participating teams. Game jams (rather than singular game development teams) could also be designed to focus on development of new “features” for established free-to-play games that are taking over the world of online, mobile games. Game jams with such foci may be of great interest to the game studios, old and new, as ways to demonstrate whether/how such game jams can create MMOGs/F2P features with plausible economic value as well as enabling (student) participants experience in game development practices that are not as well addressed by traditional game design approaches. As before, whether such competitions are seen as providing asymmetric benefit to the game studios (thus denoting a playbour dynamic [13, 37]), or are also mutually beneficial to participating students/indie game developers [27, 31]), remains an open question. Fifth, as game jams vary in the duration, product/artifact submission requirements, team size, and role composition, then there are conceptual challenges for how to articulate both plausible and reusable research methods for field studies or systematic artifact studies. Similarly, how might a game jam be designed to focus the efforts of participating teams on maximizing product quality and development productivity, within limited duration jams? Such a question addresses a classic challenge in identifying how best to engineer a new software system within time and budget constraints. Sixth, the five field studies and different forms of game development competitions help reveal that such events vary by the amount of technical and creative leadership versus free choices provided by local game jam organizers. The more short-term and informal, the less guidance and commitment required, while longer-term and more structured, the greater guidance, commitment, and quality assurance oversight
Local Game Jam Software Development Organization
53
required. This is of course a lesson learned long ago within the SE community, and thus one that is shared and relearned by game jam organizers and focused game development project leaders. The competitive element of game making on a common time box schedule, and on reliance of project leaders to mentor, plan, and guide development teams, helps better prepare participants for understanding what is required for making games that must go beyond being recognized as “weekend wonders.” Accordingly, whether software engineering educators will choose to embrace longer-term challenges that can arise from repeated SE-focused game jams, or gamefocused SE capstone project courses remain an open question, except for those who recognize the academic research opportunities such innovation factories can provide [1, 17, 24]. Seventh, the five field studies also demonstrate the potential to engage in experimental SEE projects and practices. Many students (and probably most software engineering educators) can benefit from first engaging in comparative game analysis studies that employ analytical rubrics/templates for classifying and assessing the features and capabilities of different games appropriate for student engagement [24]. Such analysis can yield insights about the relative ease/difficulty of reusing or repurposing the functional capabilities observed in games open for analysis. Insights also come from actually playing the games being analyzed—again, this might be an uncertain challenge for established software engineering educators, rather than for students. Furthermore, insights also come from classroom and homework efforts that focus attention to designing and playtesting of analog games—games that do not involve computers or programming—that help produce a better sense of how game mechanics might make a new game fun and playful, Finally, from more of an academic research perspective, the time box limited schedules of intramural game jams and capstone project coursework allows for both natural and controlled experiments when involving enough students that can be formed into comparable team groups [2, 3]. Similarly, the comparative analysis of games conforming to a theoretical sample or game genre of interest, affords the opportunity to engage in domain analysis that in turn can give rise to reusable game software patterns that can shape future game research, development, and education [31]. Eighth, game jams traditionally ignore or marginalize topics that are important to commercial game software success. These include: (a) how to support secure cheating-free game play, as well as utilize secure in-game (or in-app) content purchase interfaces; (b) how to develop games that resist efforts of players to cheat, or create grief for unsuspecting players; (c) how to embrace the rise of game architectures for free-to-play games that are centered about periodic addition of modular “features” (new in-game play objects or character adornments rather than expansion packs) that support game play update experiences and micro-transactions for purchasing in-game (or across game) resources/assets; (d) game jams that encourage user-directed extension or repurposing other existing games that are open to such evolutionary adaptation [27, 30, 31]; (e) how to make serious game jams more open to participants with domain-specific expertise who may not be skilled in game design, programming, or artwork creation, but who can help specify valuable domain topics for players to learn or master; (f) games produced with built-in libraries and architectural middleware providing scalable support for multiple players; and (g) how best to rapidly produce
54
Software Engineering Perspectives in Computer Game Development
games within a jam style competition that embrace software sustainability issues that do arise (e.g., software designed for reuse as a sustainability technique), or do not arise (games that are not envisioned as products, but instead as demonstrations of mastery or development skill potential) in game development competitions. Last, the emergence and pervasiveness of mobile games is also transforming the landscape for game production and dissemination. Games have become the most popular kind of software app found in online mobile app stores like Apple App Store and Google Play, or via online game publishing networks like Steam for desktop computer games. Thus, there is also opportunity for structuring game jams around the utilization of commercial game publishing platform services such as these, as games so published are games that student developers can highlight as career milestone products. Developing games for these platforms can entail from simple to complex integration of platform libraries accessed through application programming interfaces (APIs) that can provide statistics on downloads and usage patterns, elicit remote user feedback and review ratings, accommodate game sales for confident game developers, and more.
2.5 CONCLUSIONS Game software development competitions are fun, hard work, low-cost, of varying duration, and intensive. Most are neither motivated nor rewarded academically (i.e., no transcript grades or formal examinations given). However, they can be intra-mural or inter-mural, and can stipulate balanced or ad hoc team configurations. Capstone game projects can also be organized as intramural competitions that may or may not have external game project sponsors. In such capstone project courses, emphasis is generally focused on learning how to practice and demonstrate competency in game software development, but in the future may also focus on CGSE. Game competitions can also be open-ended or closed-ended. But their organizations often at times impose constraints that may unnecessarily inhibit the competitors, limit their skill development or learning, impose game development frameworks that limit rather than facilitate creativity and rapid development. Multi-team game development competitions vary by the amount of leadership, commitment, and guidance provided to participants by the competition organizers. Lower effort requires more independent choice by game makers, greater risk of making common/known mistakes, and good luck, while higher effort reduces choices, requires more resources and project management, may produce more sophisticated and accomplished results, yet does not guarantee a winning product. Such are the lessons so far seen within CGSE as a lens focusing on game jams and related game making competitions. Game software development competitions can serve as a test bed for exploring, observing, or evaluating new SE tools, techniques, and concepts. Such events can therefore also serve as field sites for careful empirically grounded field studies of game making processes and practices. Equalized and balanced competitions represent time-compressed ways and means for conducting empirical SE studies. Competitions can precede or follow SEE coursework, as follows: jams conducted before SEE coursework draw attention to raw talent, while jams after SEE
Local Game Jam Software Development Organization
55
coursework emphasize demonstration of learned SE skills that is more accomplished and better balanced, if the SEE was successful. This may help students and others in industry understand the value of presenting SE experiences in ways that entail tough technical, time-constrained team collaboration challenges that are ultimately perceived as a fun thing to do.
ACKNOWLEDGMENTS Preparation of this article was supported by grant #1256593 from the National Science Foundation. No review, approval, or endorsement is implied.
REFERENCES [1] S. Beecham, T. Clear, D. Damien, J. Barr, J. Noll and W. Scacchi (2017). How Best to Teach Global Software Engineering, IEEE Software, 35(1), 16–19. [2] S. Bendifallah and W. Scacchi. (1989). Work Structures and Shifts: An Empirical Analysis of Software Specification Teamwork, Proc. 11th. Intern. Conf. Software Engineering, Pittsburgh, PA, ACM Press, 260–270. [3] B. Boehm, T. Gray, and T. Seewaldt. (1984). Prototyping Versus Specifying: A Multiproject Experiment. IEEE Trans. Software Engineering, 10(3): 290–303. [4] D. Callele, E. Neufeld, and K. Schneider. (2005). Requirements Engineering and the Creative Process in the Video Game Industry, Proc. 13th Intern. Conf. Requirements Engineering, (RE'05), Paris, France. 240–250. [5] K. Claypool and M. Claypool. (2005). Teaching Software Engineering through Game Design, in Proc. 10th SIGCSE Conf. Innovation and Technology in Computer Science Education (ITiCSE '05), Portugal. pp. 123–127. [6] K.M. Cooper. and C. Longstreet. (2015) Model-Driven Engineering of Serious Games: Integrating Learning Objectives for Subject Specific Topics and Transferable Skills, in K.M. Cooper. and W. Scacchi. (Eds.), Computer Games and Software Engineering, CRC Press, Taylor & Francis Group, Boca Raton, FL, 59–90. [7] A. Dorling and F. McCaffery. (2012). The Gamification of SPICE, in Software Process Improvement and Capability Determination, Communications in Computer and Information Science, Berlin, Germany, Volume 290, Springer, pp. 295–301. [8] A. Dunne and F. Raby (2013). Speculative Everything: Design Fiction and Social Dreaming, MIT Press, Cambridge, MA. [9] T. Francis (2012). The Indies Guide to Game Making, PC Gamer UK, 246. November. [10] A. A. Grossman. (2003). Postmortems from Game Developer: Insights from the Developers of Unreal Tournament, Black and White, Age of Empires, and Other TopSelling Games. Focal Press, San Francisco, USA. [11] I. Hamilton. (2011). UCI Students Build Games in a Week, OC Register, http://ocunwired.ocregister.com/2011/04/13/uci-students-build-games-in-a-week/7131/ Accessed 15 April 2015. [12] C. Kaitila. (2012). The Game Jam Survival Guide, Packt Publishing, Birmingham, UK. [13] J. Kücklich. (2005). Precarious Playbour: Modders and the digital games industry. Fiberculture Journal, 3. http://www.journal.fibreculture.org/issue5/kucklich.html. [14] M. Lampolski and W. Scacchi (2016). Learning Game Design and Software Engineering through a Game Prototyping Experience: A Case Study, in Proc. 5th Games and Software Engineering Workshop (GAS 2016), 38th Intern. Conf. Software Engineering, ACM, Austin, TX.
56
Software Engineering Perspectives in Computer Game Development
[15] I. Mistrík, J. Grundy, A. van der Hoek, and J. Whitehead. (2010), Collaborative Software Engineering, Springer, New York. [16] E. Murphy-Hill, T. Zimmerman, and N. Nagappan. (2014). Cowboys, Ankle Sprains, and Keepers of Quality: How is Video Game Development Different from Software Development?, in Proc. 36th Intern. Conf. Software Engineering (ICSE 2014), ACM, Hyderabad, India, 1–11, June. [17] J. Musil, A. Schweda, D. Winkler, and S. Biffl. (2010). Synthesized Essence: What Game Jams Teach about Prototyping of New Software Products, in Proc. 32nd Intern. Conf. Software Engineering (ICSE'10), ACM, Cape Town, SA, 183–186. [18] E. Navarro and A. van der Hoek. (2004). Software Process Modeling for an Educational Software Engineering Simulation Game, Software Process Improvement and Practice, 10 (3), 311–325. [19] M. Newland. (1984) Marv Newland's Anijam, International Rocketship Limited, https:// www.youtube.com/watch?v=4DMopYl8eR4, Accessed 28 April 2018. [20] O.A. O’Hagan, G. Coleman, R.V. O’Connor. (2014) Software Development Processes for Games: A Systematic Literature Review. In B. Barafort, R.V. O’Connor, A. Poth, R. Messnarz (eds.) Systems, Software and Services Process Improvement. EuroSPI 2014. Communications in Computer and Information Science, vol 425. Springer, Berlin, Heidelberg. [21] J. Pirker, A. Kultima, and C. Gütl. (2016). The Value of Game Prototyping Projects for Students and Industry. Proc. Intern. Conf. Game Jams, Hackathons, and Game Creation Events (GJH&GC '16). ACM, New York, NY, USA, 54–57. DOI: https://doi. org/10.1145/2897167.2897180 [22] J.A. Preston, J. Chastine, C. O’Donnell, T. Tseng, and B. MacIntyre. (2012). Game Jams: Community, motivations, and learning among jammers. Intern. J. Game-Based Learning, 2(3), 51–70. [23] L. Reng, H. Schoenau-Fog, and L.B. Kofoed. (2013). The Motivational Power of Game Communities - Engaged through Game Jamming, Proc. Foundations of Digital Games 2013 Workshop on the Global Game Jam, Chania, Crete, Greece, 2013. [24] W. Scacchi. (1991). The Software Infrastructure for a Distributed System Factory, Software Engineering Journal, 6(5), 355–369, September. [25] W. Scacchi. (2004). Free/Open Source Software Development Practices in the Game Community, IEEE Software, 21(1), 59–67, January/February. [26] W. Scacchi. (2010). Collaboration Practices and Affordances in Free/Open Source Software Development, in I. Mistrík, J. Grundy, A. van der Hoek, and J. Whitehead, (Eds.), Collaborative Software Engineering, Springer, New York, 307–328. [27] W. Scacchi. (2010). Computer Game Mods, Modders, Modding, and the Mod Scene, First Monday, 15(5), May. [28] W. Scacchi. (2011). Modding as an Open Source Software Approach to Extending Computer Game Systems, Intern. J. Open Source Software and Processes, 3(3), 36–47, July-September 2011. [29] W. Scacchi. (2012). Competitive Game Development: Software Engineering as a Team Sport. Keynote Address, 2nd Intern. Workshop on Games and Software Engineering (GAS2012), Intern. Conf. Software Engineering, Zurich, CH, May 2012. [30] W. Scacchi. (2017). Practices and Technologies for Computer Game Software Engineering, IEEE Software, 35(1), 110–116, January-February 2017. [31] W. Scacchi. and K.M. Cooper. (2015). Emerging Research Challenges in Computer Games and Software Engineering, in K.M. Cooper. and W. Scacchi. (Eds.), Computer Games and Software Engineering, CRC Press, Taylor & Francis Group, Boca Raton, FL, 261–283.
Local Game Jam Software Development Organization
57
[32] J. Schell (2014). The Art of Game Design: A book of Lenses, 2nd Edition, AK Peters/ CRC Press, Burlington, MA. [33] L. Sheldon. (2011). The Multiplayer Classroom: Designing Coursework as a Game, Cengage Learning PTR, Independence, KY. [34] K. Shin, K. Kaneko, M. Matsui, et al. (2012). Localizing Global Game Jam: Designing Game Development for Collaborative Learning in the Social Context, in A. Nijholt, T. Romano, and D. Reidsma. (Eds). Advances in Computer Entertainment, Lecture Notes in Computer Science, Vol. 7624, Springer, Berlin, 117–132. [35] J. Tanenbaum, K. Tanenbaum, K. Ibestier, K. Abe, A. Sullivan, and L. Anzivino (2015). Costumes & Wearables as Game Controllers, TEI '15 Proc. Ninth International Conference on Tangible, Embedded, and Embodied Interaction, Stanford, CA. 477–480. [36] T.L. Taylor. (2013). E-Sports and the Professionalization of Computer Gaming, MIT Press, Cambridge, MA. [37] T.L. Taylor, K. Bergstrom, J. Jenson, and S. de Castell. (2015). Alienated Playbour: Relations of Production in EVE Online, Games and Culture, doi:10.1177/1555412014565507 [38] V. Verigames (2015). http://www.verigames.com/, Accessed April 2018. [39] A.I. Wang. (2011). Extensive Evaluation of Using a Game Project in a Software Architecture Course, ACM Transactions on Computing Education, 11(1), 1–28. [40] A.I. Wang. (2015). The Use of Game Development in Computer Science and Software Engineering Education, in K.M. Cooper. and W. Scacchi. (Eds.), Computer Games and Software Engineering, CRC Press, Taylor & Francis Group, Boca Raton, FL, 31–58. [41] Wikipedia (2018), Hackathon, http://en.wikipedia.org/wiki/Hackathon. Accessed 15 May 2018. [42] A. Zook, and M.O. Riedl. (2013). Game Conceptualization and Development Processes in the Global Game Jam, Proc. Foundations of Digital Games Workshop on the Global Game Jam, Crete, GR.Received May 2018.
3
Developing Essential Software Engineering Competencies Supported by Digital Game-Based Learning Jöran Pieper Hochschule Stralsund – University Of Applied Sciences, IACS, Germany
CONTENTS 3.1 Introduction����������������������������������������������������������������������������������������������������� 60 3.1.1 Research Objectives���������������������������������������������������������������������������� 62 3.1.2 Research Design���������������������������������������������������������������������������������� 62 3.1.3 Outline������������������������������������������������������������������������������������������������� 62 3.2 Kernel and Language for Software Engineering Methods (Essence)������������� 63 3.2.1 The Essence Kernel����������������������������������������������������������������������������� 63 3.2.2 Alphas, Alpha States, and Their Checklists���������������������������������������� 64 3.2.3 Activity Spaces—The Things to Do��������������������������������������������������� 65 3.2.4 Utilizing Essence to Drive an SE Endeavor���������������������������������������� 65 3.2.5 Essence and Traditional Methods and Improvement Frameworks����� 66 3.2.6 Conclusions����������������������������������������������������������������������������������������� 66 3.3 Exploring the Field of Game-Based Learning in SE Methods Education�������������������������������������������������������������������������������� 66 3.3.1 Review Design������������������������������������������������������������������������������������ 67 3.3.2 Results and Conclusions��������������������������������������������������������������������� 67 3.4 Developing Games and Tools within Context and under Constraints������������� 70 3.4.1 Essence Kernel Puzzler����������������������������������������������������������������������� 71 3.4.2 Essence Kernel Navigator������������������������������������������������������������������� 72 3.4.3 Essence Simulation Game������������������������������������������������������������������� 74 3.4.4 Focusing on the Essentials and Supporting Knowledge Transfer���������������������������������������������������������������������������� 74 3.4.5 Facilitating Interaction and Discussion in the Real World������������������ 75 59
60
Software Engineering Perspectives in Computer Game Development
3.4.6 Supporting the Acquisition of New Knowledge���������������������������������75 3.4.7 Gameplay�������������������������������������������������������������������������������������������� 75 3.4.8 Teamwork—Collaboration and Competition�������������������������������������� 76 3.4.9 Rewarding Good SE Practice at Scoring and Ranking����������������������� 77 3.4.10 No Levels and Game Badges�������������������������������������������������������������� 78 3.5 Evaluating Study��������������������������������������������������������������������������������������������� 78 3.5.1 Study Design��������������������������������������������������������������������������������������� 78 3.5.2 Characteristics of Participating Groups���������������������������������������������� 78 3.5.3 The Procedure of Group 1������������������������������������������������������������������� 79 3.5.4 The Procedure of Group 2������������������������������������������������������������������� 80 3.5.5 Why is There No Comparative Experiment?�������������������������������������� 80 3.5.6 Why Are There No Pre-Test/Post-Tests?��������������������������������������������80 3.5.7 Questionnaire Design and Statistical Hypotheses Testing������������������ 81 3.6 Results.............................................................................................................. 81 3.6.1 In-Game Measurements of the Essence Kernel Puzzler��������������������� 81 3.6.2 In-Game Measurements of the Simulation Game������������������������������� 81 3.6.3 Questionnaire Results������������������������������������������������������������������������� 82 3.6.4 Course Project Work��������������������������������������������������������������������������� 84 3.7 Discussion and Conclusions���������������������������������������������������������������������������� 84 3.8 Future Work and Research Directions������������������������������������������������������������ 86 Acknowledgments���������������������������������������������������������������������������������������������������� 86 Notes������������������������������������������������������������������������������������������������������������������������ 86 References���������������������������������������������������������������������������������������������������������������� 87
3.1 INTRODUCTION Software Engineering (SE) education is challenged in many ways. Educators and curriculum designers must take enormous diversity into account and offer a wellrounded range of knowledge to enable tomorrow’s software engineers. Conditioned by limited time, all contents provided are just a selection out of a vast, diverse spectrum. While depth and breadth of selected contents vary in detail, it is of consent, that beside all technical aspects and tools a profound knowledge of SE methods, which describe approaches to the production and evolution of software, are crucial for successful SE practice. Emerging agile and lean approaches propagate the renouncement of ballast—the disuse of practices or tools, which form the foundation of more traditional development processes. Lacking experience in developing complex software systems, the consequences of (not) using practices and tools can be judged hardly. While some of the SE knowledge areas are well suited to be learned by lectures and associated laboratory tutorials, others are less eligible. Those include SE methods1. “These are critical topics for industrial practice, yet it is a particular challenge to motivate students to feel passionate in these areas, and hence learn what they need to know.” [2] Without experience building more complex software systems the need for such disciplined approaches may not be as apparent as desired. A perception that (trivial) projects in the past were manageable without “restricting rules” is widely spread. Discussion, deeper analysis, or comparison of SE practices and methods in
Developing Essential Software Engineering Competencies Supported
61
the form of a dialog, which would be beneficial to learning, is complicated by the fact that students often cannot draw on any own experiences. So, they would have to take it for granted—or not. In course or capstone projects, commonly required by curriculum guidelines, students losing a holistic view due to remarkable cognitive load can be observed. Just delivering artifacts under time pressure, requested by a preselected software process, can hardly provide the deep impression that practices, methods, and tools were supporting students’ work. In the course or capstone projects the following observations can be made: • Students struggle to orientate inside a given SE method. • They have got difficulties answering the question(s) by whom, how, when, and in particular why specific activities should be accomplished. • The reasonable division of responsibilities and a rigid fixation on technological and functional details make them lose the holistic view on the development process as a whole—with all its relevant dimensions. • Takeaways from such projects are too specific and hardly transferable to other contexts and future challenges. In his review of SE in the 20th and 21st century, Boehm concludes that SE education is required to “Keeping courses […] up-to-date; Anticipating future trends and preparing students to deal with them; Monitoring current principles and practices and separating timeless principles from out-of-date practices; Packaging smaller-scale educational experiences in ways that apply to large-scale projects; [… and] Helping students learn how to learn, through […] future-oriented educational games and exercises […].” [3] The binding of current approaches to single software processes, both in conducting course projects as well as in game-based approaches, limits their transferability to new contexts that require a different set of SE practices. Sommerville emphasized that “engineering is all about selecting the most appropriate method for a set of circumstances.” [4] To enable conscious decisions of tomorrow’s software engineers, they have to know their options and be able to take multiple aspects and perspectives into account. Preceding research showed that simulation and games were able to foster empathy with the necessity of employing software processes to accomplish more complex software projects to a certain level. However, evaluation of these approaches showed that they were not able to impart new knowledge too. [5] At this point, SE education has to provide concepts for retaining a holistic view of the whole SE endeavor while working focused on the required details at the same time. SE education must provide competencies—clusters of related knowledge, skills, and attitudes enabling persons to accomplish all the tasks in a given context. Moreover, these competencies must also be transferable to different contexts to meet students’ upcoming challenges. In the field of software processes, a trend toward flexible practices and composed of SE methods, tailored to the needs of a team, instead of monolithic software processes is noticeable. The proponents deliver strong arguments to do so. SE education is asked to utilize these advantages to achieve its educational objectives.
62
Software Engineering Perspectives in Computer Game Development
3.1.1 Research Objectives This research aims at overcoming the described challenges. It is in search for a DGBL approach that provides students with a familiar environment already appreciated at the start of the course or capstone projects, lowering the overall cognitive load, and increasing the transferability of the knowledge gained. Instead of offering a somewhat isolated game-based learning activity, this research aims at delivering an approach to provide transferable skills preparing students for their real project, work to enable experiences that support the development of desired SE attitudes. All in all, students should be empowered to apply a systematic, disciplined, and quantifiable approach to the development of their software—and appreciate the guidance given by SE concepts and methods. “Each meaning, which I realize by myself, each rule, I establish according to my own lights, convinces me more and motivates me higher than any extrinsic meaning that I hardly comprehend.” [6] This guideline of constructivist didactics describes an ideal result of learning and deep understanding. To foster such learning, this research targets an active constructivist learning experience utilizing DGBL that is encouraging intrinsic motivation and social interaction through articulation, reflection, collaboration, and—where applicable—competition.
3.1.2 Research Design Based on a broad literature review about required SE competencies, learning theories, and existing game-based approaches in the field of SE methods, requirements were defined for a DGBL approach that allows students to acquire essential and highly transferable SE competencies. Based on a comparison of existing meta-models for standardized software process description concerning their suitability in the context of SE education, SEMAT Essence was chosen as the foundation of a gamebased approach. An intrinsic and reflective game, tightly linked2 to the underlying domain and addressing higher competency levels, as well as a supporting game and an integrating tool to be utilized in that approach had to be designed, implemented, and integrated to provide a learning experience that got evaluated in a case study.
3.1.3 Outline This chapter is structured as follows: Section 2 provides some background about SEMAT Essence, the OMG standard providing the foundation of games developed to achieve the objectives described. Section 3 summarizes results from a literature review that provided findings on state of the art in game-based learning in SE education and helped to develop suitable design strategies for the games that got developed. These games and underlying concepts are described in detail in Section 4. The evaluation of concepts and approaches taken was conducted with a case study that gets described in Section 5. This chapter closes with a discussion of results in Section 6 and delivers an outlook of future work and proposed research directions in Section 7.
Developing Essential Software Engineering Competencies Supported
63
3.2 KERNEL AND LANGUAGE FOR SOFTWARE ENGINEERING METHODS (ESSENCE) SEMAT3 Essence, an emerging OMG standard [9], provides a language for the unified description of SE practices and methods as well as a compact kernel, capturing the essentials of any SE endeavor in a process- and practice-independent way. The kernel provides the foundation to create practices, which eventually get picked and composed to complete SE methods, tailored to the needs of SE teams in their respective contexts. Designed primarily for software professionals’ daily practical and intuitive use, it delivers a significant paradigm shift compared to other attempts of unified process definition, for example, SPEM or ISO24744, which are primarily targeting process engineers. Instead of just describing an SE method, the actionable kernel provides support to use a method dynamically while working on an SE endeavor. This section describes some of the characteristics that make Essence a good foundation for this research, both, regarding the support of objectives of SE education as well as providing a foundation for educational games.
3.2.1 The Essence Kernel The Essence Kernel “is a stripped-down, light-weight set of definitions that captures the essence of effective, scalable software engineering in a practice independent way.” [9] It was designed to be as small and universal as possible to represent the inherent essence of each SE endeavor. The Essence authors separated the common ground, a stable kernel, from the varying details, depending on the respective context represented in SE practices. With that common foundation, practices can be defined, exchanged, and applied independently. They can be cherry-picked, mixed, and matched by project teams, organizations, and communities based on their unique needs. By combining the kernel and a set of chosen best (fitting to the context) practices, teams are empowered to form their own methods. This method composition is illustrated in Figure 3.1. Providing a common ground, an anchor and thinking framework, this kernel facilitates learning of new SE practices and methods—making it easier to notice similarities and differences between those newly introduced practices and those already known by a person or organization. To provide the essential elements of any SE endeavor, the Essence Kernel uses the concepts of Alphas, Activity Spaces, and Competencies. To facilitate orientation inside of the kernel, it is structured into three discrete areas of concern, each focusing on specific aspects of SE. Table 3.1 summarizes the elements of the kernel. Interrelationships between the seven provided Alphas and between Alphas, their states, and Activity Spaces are defined. The kernel omits the inclusion of other elements, like work products and activities, to preserve its independence from any practice. These elements are reserved for practices, which link their elements to the kernel’s respective elements. The Essence Kernel is made tangible by providing a card metaphor. SEMAT provides a set of printable cards containing Alphas’ definitions and Alpha States’
64
Software Engineering Perspectives in Computer Game Development
+
= Team’s Method
Essence Kernel
+
Requirements Elicitaon Pracce (e.g. Use Cases)
+ Acceptance Tesng Pracce
+ Mulple Pracces from Various Sources (e.g. Connuous Delivery)
FIGURE 3.1 SEMAT Essence Method Composition by Combining a Kernel and Chosen Practices.
TABLE 3.1 Essence Kernel Elements Area of Concern
Alphas
Activity Spaces
Competencies
Customer
Stakeholders, Opportunity
Stakeholder Representation
Solution
Requirements, Software System
Endeavor
Team, Work, Way-of-Working
Explore Possibilities, Understand Stakeholder Needs, Ensure Stakeholder Satisfaction Understand the Requirements, Shape the System, Implement the System, Test the System, Deploy the System, Operate the System Prepare to do the Work, Coordinate Activity, Support the Team, Track Progress, Stop the Work
Analysis, Development, Testing
Leadership, Management
checklists. Such a set of cards is easy to carry around and supports agile team sessions. By summarizing the most important concepts, cards facilitate learning by doing without the need for rote memorization before starting to use Essence. Figure 3.2 shows an example of both, an Alpha card and an Alpha State card.
3.2.2 Alphas, Alpha States, and Their Checklists Alphas represent the essential elements of an SE endeavor that are relevant to a continual assessment of the progress and health of an endeavor. [9] Each Alpha is provided with a set of Alpha States. Each Alpha State owns a checklist supporting its assessment (cf. Figure 3.2). An SE endeavor is progressed by progressing the states of all Alphas and retaining their health. The checklist provided to assess each of the Alpha States provides immediately available expert knowledge enabling even less experienced teams to think holistically about the endeavor and to recognize aspects that otherwise, based solely on team members’ experiences, might be forgotten.
Developing Essential Software Engineering Competencies Supported
65
FIGURE 3.2 Examples of an Alpha Card and an Alpha State Card for the Alpha “Way-ofWorking” [10].
3.2.3 Activity Spaces—The Things to Do A team must perform an activity addressing the next target state to progress an Alpha. The Essence Kernel does not provide concrete activities since they would prescribe a specific way to perform a set of tasks and bind the kernel to a particular practice. Instead, the Essence Kernel provides a set of Activity Spaces serving as abstract placeholders. These Activity Spaces are the things to do in every SE endeavor. Activity Spaces are provided with entry and completion criteria represented by Alpha States. While Activity Spaces provide the What and (concerning the Alpha States) When to do, practices deliver concrete activities—the How to do. Activities provided by an Essence practice are grouped and classified by an Activity Space.
3.2.4 Utilizing Essence to Drive an SE Endeavor Essence and its kernel support a structured, quantifiable, and goal-oriented approach to progress an SE endeavor. A typical proceeding follows a Plan-Do-Check-Adapt (PDCA) cycle4 [12]: 1) Plan: A team starts with assessing the current state of the endeavor. Alphas and their Alpha States ensure that all relevant dimensions are considered. Checkpoints support the assessment of each of the Alpha States. Walkthroughs or Assessment Poker sessions are suitable to assess an endeavor in a team session. [12] The Activity Spaces addressing next targeted Alpha States provide support by identifying next activities. Checkpoints of next planned Alpha States help to identify necessary tasks. 2) Do: The team works on accomplishing the identified tasks. Emerging obstacles are removed as they occur. 3) Check: By continuously tracking progress toward defined objectives and tasks, the team ensures to keep focused and following its chosen way of working.
66
Software Engineering Perspectives in Computer Game Development
4) Adapt: Team’s way of working is continuously reviewed and assessed in the same way as all other essential dimensions (Alphas) of the endeavor. As a team recognizes impediments, caused by their way of working, better (suiting) ways of getting things done get identified and used. Plans are adapted accordingly.
3.2.5 Essence and Traditional Methods and Improvement Frameworks While the Essence language can be used to implement existing SE methods, it “is not in competition with any existing practice or method. It is agnostic to your chosen approach and can be implemented by your team today without changing what you are currently doing. It is more about helping you do what you are already doing so you can keep doing it even better in the future.” [13] By adding a common ground, or a reference model, that all teams can use, regardless of their chosen (agile or not) method, Essence adds to the ability to continually measure health and progress of endeavors and to improve teams’ way of working. [14, 15]
3.2.6 Conclusions Essence provides a shift toward flexible and composable SE practices instead of monolithic software processes prevalent in past and presence. With the Essence Kernel, this standard provides a compact, practical, actionable, and extensible set of elements inherent in all SE endeavors. As such this kernel serves as a thinking framework as well as a foundation to define practices and compose them to tailored methods. Since the kernel is actionable, it enables a team to continually and holistically monitor progress and health of all essential dimensions efficiently. With utilizing a PDCA cycle to steer the SE endeavor, the kernel provides the foundation for continuous improvement throughout an SE endeavor. A mapping of Essence’s characteristics with competencies demanded by SE curriculum guidelines manifests its suitability in SE education. As Ng and Huang, experimenting with Essence concepts in a workshop with professors and Ph.D. students, concluded “Universities cannot teach everything that industry requires. But university can provide students with a firm grasp of the fundamentals and give them the tools to learn and understand the diversity of software engineering later in their career.” [16] Therefore, SEMAT Essence was chosen as the foundation of the approaches taken in this research.
3.3 EXPLORING THE FIELD OF GAME-BASED LEARNING IN SE METHODS EDUCATION To consider the findings of relevant preceding research, the field of game-based learning approaches in SE education was explored.
67
Developing Essential Software Engineering Competencies Supported
3.3.1 Review Design Several rigorous and comprehensive meta-studies [5, 17–19] provided a good starting point to investigate the field of game-based learning activities in SE education. The focus of those studies varied but delivered an excellent foundation for data extraction of aspects of interest for this research. An own SLR to catch up with most recent developments was conducted. The research design of that SLR widely follows the procedures of [5] but includes the years 2014–2016. Additionally, doctoral theses related to the field and a paper [20], which became known to the author at a national conference and was therefore not covered by the search criteria, were considered too. [21]
3.3.2 Results and Conclusions Table 3.2 summarizes characteristics of the Systematic Literature Reviews (SLRs). These results of the literature reviews naturally overlap to some extent. The timeline of DGBL approaches in SE education proves an ongoing interest and research activity in this field. [21] Recurring entries of simulation games indicate the effort affiliated with the development and evaluation of approaches—it may take several years from initial concept descriptions till the implementation and evaluation of approaches. Some implemented games or environments, i.e., SimSE, SESAM, and AMEISE, show an impressive body of work of several researchers covering long periods, both for refinement of initial versions as well as the evaluation of taken approaches. Besides the utilization of DGBL, pure non-game-based rather industrial simulation models were used in training and education sessions. Several attempts to utilize DGBL in SE education have been made. Several games were focused on specific, somewhat isolated, SE knowledge areas. Table 3.3 summarizes such games that are addressing the area of SE methods that are of particular interest for this research.
TABLE 3.2 Systematic Literature Reviews of Simulation and DGBL in SE Education Period Simulation games Only computer-based Non-Game simulations Number of papers Number of simulators and games
[6]
[52]
[37]
[17]
This Research[19]
1996-2006
1990-2008
1999-2009
open-2013
2014-2016
Y Y
Y N
Y Y
Y Y
Y N
N
Y
N
Y
N
21 12
16 12
16 8
42 15
17 13
68
TABLE 3.3 Existing Simulation Games in SE Methods and Management Education and Training SimjavaSP [46]
SESAM [9, 10, 24], AMEISE[26]
90 Incredible Manager [2]
90 OSS [45]
90 MO-SEProcess [53]
90 SimVBSE [16]
90 Simsoft [5]
90 Nassal’ s PMG [28–30]
90 SPIAL [38]
90 PMG-2D [23]
S
S
S
S
S
M
S
M
S
S
S
-
-
-
-
-
-
-
+a
-
-
-
-
-
ob
-
-
-
-
-
-
-
-
+
+
+
+
o
o
+
+
+
+
+
-
-
-
-
+
+
-
-
-
o
-
+
-
+
-
-
-
-
-
+
+
-
6
1
3c
1
1
1
1
1d
1
1
1e
-
-
-
-
-
-
-
-
-
of
og
-
-
-
-
-
-
-
-
+
-
-
“+” = yes/included, “-” = no/not included, and “o” = partially included a Players are collaborating in teams. No competition. b Only for administrative purposes. c Two models are provided by the AMEISE project. d “Work-to-do, review, rework, work-completed cycle.” e “Life cycle of the project management process according to PMBOK.” f With extracted rules from CMMI. g Based on and linked with PMBOK.
Software Engineering Perspectives in Computer Game Development
Single-player (S) or multi-player (M) game Collaboration and/or competition of players Team, course and lecturer dashboards Player in the role of a project manager Player(s) with different roles or tasks Enabled for different SE processes/methods Number of available predefined SE process/ method models Integrating industry standards for software process description Utilizing real world tools, used in SE contexts
SimSE [31, 32]
Developing Essential Software Engineering Competencies Supported
69
The educational approaches taken within these games vary considerably. Simsoft is explicitly oriented toward Problem-Based Learning. All approaches represent more or less constructivist learning approaches by providing kinds of experiential learning, active learning, learning by doing, situated learning, and discovery learning. The amount of scaffolding provided differs considerably. SimjavaSP explicitly does “not guiding” the learner, which represents kind of minimally guided instruction [38] (tending toward unguided instruction). Considering Vygotsky’s Zone of Proximal Development [39], Sweller’s Cognitive Load Theory [40], and Reigeluth’s Elaboration Theory [41], this might not contribute to ideal learning experiences. Concepts of Social Constructivism seem clearly to be underemployed in existing approaches. Players of Simsoft appreciated to work in teams, to share opinions, and learn from more experienced teammates. The extent to what learners collaborated in MO-SEProcess is not explicitly described. All other approaches are provided as pure single-player games. In these games, collaboration and motivating competition between isolated participants are not promoted. The architectures of the game environments, except those of the AMEISE project, do not enable a lecturer to get a quick summary of the performance of all players in a course. Such an overview could facilitate quick, supportive individual or group interactions in the case of occurring misconceptions as well as support debriefing activities. The usage of real-world tools, as implemented in Nassal’s Project Management Game, enables the beneficial development of skills in the knowledge domain and at using the utilized tool. On the other hand, it makes the approach depending on the tool. If the chosen tool does not represent the acknowledged mindset or essential principles of a given SE method, the approach may be rejected for just that reason. For instance, it might be less acceptable to play a Scrum-based or a Kanban-based method based on Microsoft Project, since these agile and lean methods do not advocate elaborated long-term planning and are based on pull-mechanisms. Almost all authors recognize that their games alone are not sufficient learning vehicles and hence have to be complemented by other educational approaches. Game-based learning activities must be integrated into the context as a whole in order to address intended learning objectives [5]. Most approaches offer only one fixed model of a software process or similar concepts. Model customization capabilities seem crucial in today’s software engineering education taken the real diversity of software processes into account. To transfer knowledge gained in one specific SE method to new upcoming challenges requiring different practices is an exercise left to students. None of the existing approaches facilitate a perspective on the use of flexible SE practices instead of software processes that provide common ground as a thinking framework yet. Most simulation-based approaches lack integrated documentation of the process or method to follow. Without such descriptions, players are strongly dependent on prior knowledge to accomplish the game—given that the same vocab with same semantics is used. By assigning activities, represented by just a word or short word group, concepts and semantics may keep unclear. Most simulation-based games, enabling the learning of whole SE methods, promote a focus on workload
70
Software Engineering Perspectives in Computer Game Development
optimization of team members with specific skills. Given the increasing recognition of agile and lean methods based on pull-approaches, where team members choose appropriate work packages depending on their abilities, this seems not to be an essential task in all SE contexts anymore. Unfortunately, it occupies a lot of learner’s attention—that could be focused on other essential problem-solving issues. While all of the introduced approaches provide an opportunity to familiarize with the field of SE methods and their management, each with different focus and approach, they were not explicitly built to directly support students in the course or capstone project work that is demanded by curriculum guidelines. If SE students should acquire an SE mindset, genuinely appreciating the guidance and support of utilized methods and practices, they should be provided with experiences facilitating such perceptions.
3.4 DEVELOPING GAMES AND TOOLS WITHIN CONTEXT AND UNDER CONSTRAINTS Two games, the Essence Kernel Puzzler and the Essence Simulation Game, as well as an integrating tool, the Essence Kernel Navigator, were developed. Both games and the Navigator get introduced in the following section. At designing and developing these games, clear learning objectives guided the process. The objective to support a holistic application of a systematic, disciplined, and quantifiable approach to the development of their software in course projects demands for higher learning objectives—to apply conceptual and procedural knowledge. This requires prior mastery of remembering and understanding of factual and, to some extent, conceptual knowledge5. It may seem tempting to build the one game that solves all problems at once. However, this can quickly lead to an overloaded gaming experience and require a more expensive game development. With limited resources, it seems more advisable to keep a clear focus and scope, for example, to provide learning experiences at intended competency levels. Within this research work, two games were developed. The first game addressing lower competency levels was developed to prepare a student for the second one, addressing higher competency levels and learning objectives. By separating these two games, it was possible to pick the recommended game genres to benefit from their inherent features. This is important since not all game types support all learning outcomes equally well [43, 44]. Following a methodological approach to select proper game genres [45], the genres drill and practice as well as puzzle games were selected to support the first game, the Essence Kernel Puzzler. The second game, the Simulation Game, utilizes features and characteristics of simulation and adventure games to impart the desired procedural knowledge and skills. To enable the desired learning outcomes, this was designed to be intrinsic, making the learning content an integral part of the gameplay, tightly linked to the domain, making the mastery of the learning content vital to succeed in the game, and reflective, allowing a high degree of reflection by providing as much time as needed to (re-) think about decisions.6
Developing Essential Software Engineering Competencies Supported
71
This separation also allowed for focused game design and resulting game experiences. To combine both games, they are arranged and glued by other activities in an overall learning arrangement. It is not necessary to compete with multi-million-dollar productions concerning game visuals. As Prensky states “[…] creating engagement is not about those fancy, expensive graphics but rather about ideas. Sure, today’s video games have the best graphics ever, but kids’ long-term engagement in a game depends much less on what they see than on what they do and learn. In gamer terms, ‘gameplay’ trumps ‘eye candy’ any day of the week.” [23] Learning theories provided guidance toward desired learning experiences. The games introduced in this chapter were designed with a focus on (social) constructivist approaches but without neglecting insights of cognitive learning theories arguing against minimally guided instruction. Games and simulations, which are based on models, may lead to misconceptions in learning due to reductions and “simulation shortcuts.” [16] At this point, the careful embedding of the serious game into the whole learning scenario and environment including debriefing actions is of enormous importance. “If the educators are aware of these ‘flaws’, they can use these as links to address questions that are posed or left open by and in these games. They serve as an ideal anchor for complementing educational activities.” [9] The Simulation Game introduced in this section is focused on essentials and built on top of a simplified simulation model that intentionally ignores details represented in other games. Questions that may arise from this are a welcome trigger for discussion and social interaction in a recommended debriefing session [40]. No game engine was utilized to implement the games with the desired characteristics. Instead, a technology stack common to current web development was utilized.
3.4.1 Essence Kernel Puzzler The Essence Kernel Puzzler, from now on short Puzzler, was designed as a markedly low-threshold offer to get familiar with Essence Kernel’s vocab and some of its concepts. With just choosing a nickname everyone can use the Puzzler online7. It was designed to make people entering the Essence world familiar with definitions of terms used in the Essence Kernel vocab and to delve into the relationships of these elements in a more accessible and enjoyable way. The Puzzler focuses on Alphas and their associations. It introduces Activity Spaces as well as the Competencies defined by the Essence Kernel. After mastering the Puzzler, students should be able to name elements, to assign definitions correctly, and to establish connections between them. Mapping the learning objectives to the cognitive process and knowledge dimensions of the Revised Bloom’s Taxonomy [46], it targets lower cognitive levels, esp. the first levels factual and conceptual of the cognitive process dimension as well as remember and understand of the knowledge dimension. As such, mastering the Puzzler is a good starting point before beginning to strive for higher learning goals.
72
Software Engineering Perspectives in Computer Game Development
FIGURE 3.3 Two Screenshots of the Essence Kernel Puzzler Presenting Alpha Cards and Alpha Relationships.
The Puzzler is provided with a simple user interface. Each of the eight levels gets introduced with some basic information about the topic to master. Each level presents a task to handle via dragging and dropping elements to their correct position, for example, a name of an Alpha to its corresponding card. The time needed to master the level is measured and motivates to solve the puzzle faster once repeated. Feedback about the progress is provided by counting the number of correctly assigned elements and the number of failed attempts. Unconscious trial-and-error-acting gets not rewarded. If a player fails too often, the progress of the current level gets lost, and the level starts again. The player gets an overview of achieved results in her high score list. Figure 3.3 presents two screenshots of the UI, each showing one of the eight Puzzler levels.
3.4.2 Essence Kernel Navigator The Essence Navigator represents an elementary execution environment for the SEMAT Essence Kernel as defined by the operational semantics of the Essence specification [9]. It provides the functionality to track the overall state of the SE endeavor, to determine the current overall state of the endeavor, and to give elementary advice about next steps to do. The Navigator provides its unique features in the same way to the player in a Simulation Game as it does to the teams assessing and controlling their individual course projects. The Screenshot in Figure 3.4 illustrates the Alpha overview provided to individual players and project teams by the Navigator. Each Alpha is easily recognizable with its currently assessed state. The balance indicator at the right side of the figure gives a hint if all Alphas are progressed in a balanced way. The more this indicator forms a harmonic circle, the more balanced is the progress achieved—or better assessed—so
Developing Essential Software Engineering Competencies Supported
73
FIGURE 3.4 Screenshot of the Essence Kernel Navigator: Alphas Overview.
FIGURE 3.5 Screenshots of the Essence Kernel Navigator.
far in the project. Figure 3.5 (left) illustrates Navigator’s support at assessing the current state of an Alpha. All Alpha States are provided with their respective checklists. In the example given, the Activity Space Ensure Stakeholder Satisfaction has to be performed to achieve the state Satisfied for Deployment of the Alpha Stakeholders. The Navigator integrates Activity Spaces in the learning experience and establishes
74
Software Engineering Perspectives in Computer Game Development
bidirectional links between Alpha States and the Activity Spaces leading to their achievement. Figure 3.5 (right) illustrates the detail view of the Activity Space Ensure Stakeholder Satisfaction. A short description and links to all the Alpha States that can be achieved by performing this Activity Space are provided. Based on the current assessment of all the Alphas, the Navigator provides elementary recommendations for next activities guiding a student through a virtual or a real project.
3.4.3 Essence Simulation Game The design objective of the Simulation Game is to give students a quick and profound start in their course project activities, to provide already familiar concepts and tool support—hence raise conscious goal-oriented attitude and lower the cognitive load students face in such projects. The Simulation Game addresses students already introduced to the goals of the SEMAT Essence Kernel and familiar with basic concepts like Alphas, Alpha States, their checkpoints as well as Activity Spaces. Playing the Simulation Game based on the Essence Kernel should not take longer than a standard course unit of 90 minutes and be supplemented by debriefing activities. After mastering the game, students should be able to apply acquired concepts in a simplified defined simulated context. They should feel comfortable with performing Essence’s “mechanics” and applying the PDCA-cycle. Students should be able to orientate inside a given SE method/kernel and to consider all essential dimensions (Alphas) of the SE endeavor. They should be able to assess the progress and health of their endeavor in a structured and continuous way (at least at the presented abstracted level). It would not be realistic to assume that students would know all the Alpha States and their checklists in detail after playing for about 90 minutes. However, this is not even necessary, since Essence, as described, provides cards to learn on-the-fly by doing. The game provides the opportunity to explore the multitude of different information presented to project members in an SE endeavor and to experience how Essence supports the team in structuring this information—how Essence is guiding the team through the project in a holistic way. Mapping the learning objectives to the cognitive and knowledge dimensions of the Revised Bloom’s Taxonomy, the Simulation Game targets the cognitive process dimension levels remember, understand, and apply. All levels of the knowledge dimension (factual, conceptual, procedural, and metacognitive) get addressed.
3.4.4 Focusing on the Essentials and Supporting Knowledge Transfer Since the game is designed to support students at their subsequent real project work, not to steal time from this task, the time allocated to playing the game is limited by intention. That is why the game does not impart all known phenomena and effects that might appear at driving an SE endeavor. The SEMAT Essence Kernel encloses the essential aspects common to all SE endeavors but not every detail and aspect that might be of interest only under certain circumstances. By integrating the kernel, the Simulation Game follows that approach. It is as concrete as the kernel itself. No
Developing Essential Software Engineering Competencies Supported
75
distractions were added in order to keep players focused. Topics and aspects omitted or just broached might be deepened and discussed in debriefing activities or at the following course project itself. By choosing the Essence Kernel as the foundation of the Simulation Game, students get familiar with a compact thinking framework which is not bound to any practice or software process and supports driving any SE endeavor. By integrating the fundamental mechanisms of endeavor assessment, the definition of next goals, choosing the next appropriate activities to accomplish these goals, and monitoring the progress and health of all essential dimensions directly into the gameplay, students acquire highly transferable knowledge and skills.
3.4.5 Facilitating Interaction and Discussion in the Real World The game provides occasions for discussion to facilitate thoughtful analysis and social interaction. It delivers feedback indicating how well a player’s decisions were, compared to the decisions of other team members—and how well the whole team performed, compared to other teams in the course. The game collects results of all players and their teams to support debriefing activities.
3.4.6 Supporting the Acquisition of New Knowledge In a meta-study, the authors stressed that “[…] game-based learning appears to have more impact on lower cognitive levels, reinforcing knowledge learned earlier. It seems inadequate for teaching new knowledge.” [43] While the Puzzler was designed to learn and reinforce concepts and vocab, overall addressing lower cognitive levels, it is the objective of this game to let players actively apply concepts, to analyze and reflect on their application in a PDCA-cycle.
3.4.7 Gameplay Players slip into the role of a coach directing a virtual team of three developers through an SE endeavor having a fixed budget and a fixed cost rate per day. By assigning activities to their virtual team, the endeavor gets progressed as far as all preconditions for performing the assigned activity are satisfied. Before a player may assign an activity to her virtual team, she must explore that activity first. For that, a player makes use of the Navigator integrated into the game. Inside a game environment, it provides the opportunity to explore and collect Activities. When an activity is assigned to the virtual team, it depends on the fact, if all the prerequisites to successfully perform the activity are met. If this is the case, the virtual team delivers feedback about their current tasks via messages. The list of messages grows while the game progresses and serves as an activity log, enabling the analysis of decisions taken by just scrolling down the list of messages. The analysis of a message from the virtual team might trigger a new assessment of an Alpha State. The player should utilize the provided Navigator to get support at this point. The player must find the Checkpoint corresponding to the incoming message and decide if it should be checked. The Navigator acts inside the game as it
76
Software Engineering Perspectives in Computer Game Development
would in a real SE endeavor. It does not define the state of the virtual endeavor but gives support to assess it. To a player consciously processing all incoming information, the Navigator perfectly provides the current state of his virtual endeavor—just like the Navigator would in an ideal situation in a real SE endeavor. If a player does not process the incoming information correctly, he runs into danger to make wrong decisions at assigning activities based on a misconceived impression of endeavors’ current state—just like in a real SE endeavor. When all Checkpoints of an Alpha State got fulfilled, a new Alpha State is reached. In that case, the player has to decide if a new activity should be assigned to the virtual team. The Navigator supports at this point by linking all the Activities addressing the next targeted Alpha State and by providing an elementary to-do list. Should the player assign an activity with open preconditions on its addressed Alpha States or Checkpoints, a tutor message gets sent to the player. If the player assigns only activities that cannot be progressed a tutor message gets sent, stating that no progress was made for several days. Because the cost rate per working day is fixed, a player would have spent money without achieving any progress on the endeavor and might regret decisions made. To enable learning from failure the game provides a time travel feature enabling the player to travel back in time and make decisions once again—now with increased knowledge. To avoid abuse of the time travel feature and not to promote unconscious try-error-time-travel cycles, this feature is burdened with cost, making it more attractive to try finding good decisions and to use the time travel feature only in case of need. The virtual team of a player starts with low performance to reflect the assumption that a new team starting an endeavor does not start with full performance [47]. To unfold the team’s full potential, it has to organize itself and its work, find its ideal way of working, has to get support, and track its progress. In SEMAT Essence, all these issues are implemented by the Alphas of the Endeavor area of concern. To raise the performance of the virtual team, the Alphas of this area of concern must be progressed. Figure 3.6 shows the user interface of the Simulation Game. The game provides a simple non-distracting user interface oriented toward instant messengers, requiring no further introduction. At the left side, we see the Essence Navigator, deeply integrated into the gameplay. By designing and providing a lean interface with scaffolding where needed, the extrinsic cognitive load is kept low allowing learners to focus on the essential learning content.
3.4.8 Teamwork—Collaboration and Competition Experts in the field of game-based learning demand a “wider use of social interactive learning.” [48] The game adds the idea of teams to the individual gameplay of every participant. In this setting, each player is exploring the whole game and the integrated learning content on her own but is a member of a team at the same time. Measuring the performance of the whole team as a primary success indicator of the game fosters collaboration and discussion inside the teams trying to perform better than other teams. This opens new opportunities for interaction. Utilizing provided leaderboards of team rankings as well as the ranking of individual players gives
Developing Essential Software Engineering Competencies Supported
77
FIGURE 3.6 Screenshot of the Essence Simulation Game.
orientation how optimized chosen paths are in relation to other one’s results. This orientation gives immediate feedback and is the starting point for further interactions inside a team. Since all players try to master the same challenges, a team member performing better at a given time may support other members of the team—having their common goal in mind. This approach combines collaboration aspects of social interaction inside teams with the competition aspect between teams. The game provides an elementary team chat.
3.4.9 Rewarding Good SE Practice at Scoring and Ranking As SE is defined as “the application of a systematic, disciplined, quantifiable approach to the development, operation, and maintenance of software […]” [4], the game rewards good SE practices and penalizes bad ones, assuring that the chance to get points by accident is minor. The point system integrated into the game reflects Essence’s usage in the real world and rewards conscious decisions. Essence reflects the progress of an endeavor in the Checkpoints fulfilled, and the Alpha States reached. Representing that, the game rewards fulfilled checkpoints in the underlying simulation model driven by players’ decisions. To reward conscious practice, the assessment of a Checkpoint itself gets rewarded, if that one gets checked after such an assessment is indicated by the game’s state. To facilitate players’ reflective thinking about decisions made so far, the game provides rankings and leaderboards of players inside a team, as well as a ranking of the team inside the whole course. Rankings are presented in the status bar and are always visible to the player. A click at one of the ranking stats reveals corresponding leaderboards. All players get presented with the leaderboards corresponding to their individual progressed virtual time inside the game.
78
Software Engineering Perspectives in Computer Game Development
3.4.10 No Levels and Game Badges By intention, this game does not make use of levels and badges. Instead, it makes use of the natural structure and “rewarding system” of the underlying domain. Essence provides Alphas with their Alpha States and their Checkpoints. Each time the player ticks off a Checkpoint (and that action was indicated) she gets rewarded by points and the Navigator visualizes the progress. This way, a player earns the same “badges” in the game as she does as a software engineer at driving a real SE endeavor. Reaching an Alpha State might be seen as completing a level-like stage of the game. The difficulty of the game increases naturally, driven by the inherent characteristics of the domain. The number of activities that can be (reasonably) assigned to the virtual team grows while the game is progressing. By assigning multiple activities to the virtual team at once, the number and variety of information flowing in are growing, making the proper processing and analysis of the current state of the endeavor more challenging. The decreasing performance of the virtual team, caused by the insufficient attention paid to the Alphas of the Endeavor area of concern, might add new challenges too. With these characteristics, this approach provides an intrinsic, tightly integrated, reflected, and asynchronous (turn-based) game that is utilizing a blended paradigm [7, 49].
3.5 EVALUATING STUDY This section describes the evaluation of concepts and approaches described in this chapter. A case study was conducted to evaluate concepts and games developed. Details follow in the next subsection. Also, in order to establish the consideration of all relevant dimensions, the characteristics of the approaches and the learning games provided were mapped to quality criteria of learning games proposed by Dondi and Moretti [45]. This mapping ensures that all relevant dimensions of learning games were considered, and the demanded quality criteria are provided. Two additional mappings to heuristics were conducted to compare the approaches taken with recommendations based on the synthesis of a comprehensive SLR by Jiang et al. [18] as well as with SE education requirements based on conclusions of Boehm [3] reviewing SE in the 20th and 21st century. Further was ensured that the described approaches represent the characteristics of constructivist learning environments, which are widely accepted to be facilitating learning, collected by Jonassen [21, 50].
3.5.1 Study Design This section describes the setup of the case study used to evaluate the chosen approaches. Two groups of participants took part in the case study. Both consisted of students of the Hochschule Stralsund—University of Applied Sciences, Germany.
3.5.2 Characteristics of Participating Groups Both groups were employed in the summer semester of 2016. They provided the observations and results reported. The first group (Group 1) consisted of 12 students
Developing Essential Software Engineering Competencies Supported
79
(11 male, one female) taking the course “SMIB4500—Project Seminar Software Engineering” embedded in the 4th semester of the undergraduate curriculum of “SMIB—Applied Computer Science, Software Development and Media Informatics.” Within the scope of the course, students got introduced to SEMAT Essence for the first time. Traditionally in the course, students exercise a software project from start to end, following a given software process based on the OpenUP [41] and AgileUP [49] and heavily customized to the organizational needs of the course. This software process should not be utilized in a real SE endeavor customized that way. Students got introduced to SEMAT Essence and in particular the Essence Kernel to get provided with transferable concepts and a thinking framework being of use in any future SE endeavor. The second group (Group 2) consisted of seven volunteers, visiting the 6th semester of the undergraduate program “SMIB—Applied Computer Science, Software Development and Media Informatics” and interested in learning something about SE in general and Essence in particular. This group differs to some extent from Group 1. One year before, they got shortly introduced to SEMAT Essence. Generally, students of SMIB in sixth semester already had their practical semester where they gain experience in real software industry. In their 6th semester, they are processing a bigger software project within the scope of a course. Within that course, they got in touch with Essence again. So, they were already working with Essence Kernel Cards to some extent. Students of both groups are enjoying digital games. 53% (n = 10) play regularly and further 37% (n = 7) to a lesser extent. 47.3% of the participants play more than 5 hours per week. 68% of the participants connect learning with games. None of the participants disagreed with the statement that “games provide learning.” 42% (n = 5) of Group 1 and 86% (n = 6) of Group 2 stated to have collected practical experience in software projects outside of the curriculum.
3.5.3 The Procedure of Group 1 Within the regular course students of Group 1 were introduced into SEMAT Essence by an educational case study [8], presenting a small team of software developers at kickstart of a software project. SEMAT provides this case study. Students were invited to play the Essence Kernel Puzzler in preparation of the Simulation Game. At the next course meeting, a few days later, two course units of 90 minutes were used for playing the Simulation Game and a debriefing session. The students were already organized into four teams for their project work. These project teams were used for the game too. The author provided an introduction into the gameplay, provided technical support, answered questions regarding the game, and moderated the debriefing session. Finally, students filled in the questionnaire. After this session, students used the Essence Navigator in their project work. The tool supported project team meetings and helped teams to keep on track. The Essence Navigator provided an enacted Essence Kernel. Since the process to follow was not modeled as Essence practices forming an Essence method, students had to do some mental mapping between Essence concepts, for example, Activity Spaces and activities required by the given software process to follow. In an optional exercise, the teams were asked to map explicitly the assigned activities of the given process to Activity Spaces of the
80
Software Engineering Perspectives in Computer Game Development
Essence Kernel and the points in time, when all the Alpha States got achieved, to phases and concrete iterations of the given software process. The Alpha States that were not part of the assigned project work had to be marked. At each mapping, the students were asked to provide a short explanation.
3.5.4 The Procedure of Group 2 Students were invited by the author to take part in a Simulation Game session and (optionally) to take the Essence Kernel Puzzler in preparation of the Simulation Game. Since students already had contact with SEMAT Essence, no further preparation was provided. The group was partitioned into three teams. In one session, lasting 2 hours, the simulation game was played followed by the debriefing session. The author provided an introduction into the gameplay, provided technical support, and answered questions regarding the game. Finally, the questionnaire was filled in by the students.
3.5.5 Why is There No Comparative Experiment? Having only access to quite small groups of students for this case study, it was chosen to provide all of them with the chance to learn from new concepts and approaches and to collect as much feedback on the tools and concepts developed as possible. With just 19 participants a division into treatment and control groups would have made the results less informative and significant. Furthermore, SEMAT Essence is a new standard with growing—but at the time of conducting—limited learning material. With no prepared rich and ready to use learning material, rated competitive to the provided approach, any attempt to produce that material alongside with limited resources was judged to be in danger of getting exposed to a biased performance.
3.5.6 Why Are There No Pre-Test/Post-Tests? The skills and competencies striven for in this approach and facilitated by the games provided are to orientate inside an SE endeavor. This includes utilizing SE concepts and tools provided to act in an SE endeavor in a goal-oriented and structured way, for example, by holistically assessing the current state of an SE endeavor and determining next reasonable steps to take. Such skills and the corresponding attitudes are not, or only to a small extent, measurable in a test. Rote memorization is not a primary concept of Essence, which provides the foundation of the approach. Instead, it is providing rich as well as compact documentation, for example, in the form of physical cards to carry with you. There is no need to rote memorization of all states and their checkpoints. Essence promotes learning by doing and enables to learn and deepen knowledge on the fly—once basic concepts are understood. The Essence Kernel Puzzler, as well as the Simulation Game, provides embedded assessment, ingame measurements allowing drawing inferences from the gameplay about the learning progress to, some extent. This approach avoids disadvantages associated with examinations of learning progress separated from the gameplay reported by existing studies of DGBL [51].
Developing Essential Software Engineering Competencies Supported
81
3.5.7 Questionnaire Design and Statistical Hypotheses Testing The questionnaire was designed to survey different aspects of the approaches taken. The answer types include dichotomic choices, 5-point Likert-type scales and other rating scales resulting in ordinal scaled data, as well as free text answers to collect remarks of participants. To establish statistical significance answers undergone hypothesis testing where appropriate. Having only a small set of observations (n < 20) without any knowledge about the particular distribution of the data only a non-parametric statistical test, able to handle small sets of observations, is qualified for this study. The Sign-test fulfilling these criteria was chosen. It is a non-parametric binomial test about the median η of a population.
3.6 RESULTS This section describes results of the evaluating case study. For the sake of brevity, no representation of the mapping of presented concepts on recognized heuristics done [21] is given here.
3.6.1 In-Game Measurements of the Essence Kernel Puzzler At conducting the evaluating study 76 game instances and 788 attempts to master levels of the Puzzler were counted. 508 (64.5%) of these attempts were successful while 280 (35.5%) failed. Players spent in sum 22 hours trying to master the levels of the Puzzler. On average each level of the Puzzler was played 98.5 times (sd 58.6, median 88). Not all the game instances (n = 76) resulted in a mastered first level (n = 69). While remarkable a part of the players masters each level just once, others repeat to master the levels multiple times, which might be motivated by the wish to improve the time needed or to decrease failures made. Of the 280 failed attempts to master a level of the Puzzler, 83.9% (n = 235) were mastered afterward. Overall, the number of players mastering Puzzler’s levels is decreasing from level to level. Starting with 69 players, who mastered level 1, only 21 were mastering level 8. This decrease correlates with the assumed difficulty of single levels.
3.6.2 In-Game Measurements of the Simulation Game Teams’ scores, concerning both, points for achieving Checkpoints of Alpha States in the underlying simulation model as well as points for consciously assessing these Checkpoints, indicate that the players were able to master the gameplay, the tools provided for support, and, hence, the underlying learning content. To get these points, players had to progress their virtual endeavors and assess its Alphas consciously. Players’ scores prove that they were able to follow the PDCA cycle. Players had to utilize relationships between Essence Kernel’s elements and to show that they are already familiar with them. The results indicate different types of players and learners. Results show players aiming at performing without mistakes right from the start and others, which are
82
Software Engineering Perspectives in Computer Game Development
willing to take the given opportunities to improve results iteratively. The time travel feature that was designed to provide the opportunity to learn from failures. Overall, the players of the case study traveled 95 times in time and covered 810 days thereby. Results show that the usage of this feature varies widely and ranges from 0% to 100% of team members utilizing this feature. One team of Group 1 and one team of Group 2 provide interesting extreme manifestations of this issue. While none of the members of the Group 1 team did use the time travel feature, all players the Group 2 team used it with the highest count of performed time travels. Both teams achieved high scores—at different efficiency. Group 2, the more experienced students, were not able to outperform Group 1 in the setting of this case study. The mixed rankings of teams from Group 1 and Group 2 indicate that the Simulation Game provided the opportunity to learn for both the more and the less experienced student groups.
3.6.3 Questionnaire Results 77% (n = 10) of the respondents stated that the Puzzler prepared them well for the Simulation Game. 92% of responding students (n = 12) would recommend the Puzzler to a friend or fellow student. Students stated they found it to be a neat tool to learn the vocab fast and to get familiar with first relationships between elements. The Simulation Game was considered to be fun and students attributed learning success to it. The difficulty of the Simulation Game was perceived as “just right” by 42.1% (n = 8) of the participants. 26.3% (n = 5) found it slightly easy and 31.6% (n = 6) of the players found it slightly hard. The vast majority (78.9%, n = 15) of all students assessed the duration as “just right,” 3 players (15.8%) felt the duration was “too long” and only one player (5,3%) felt it was “too short.” Students (95%, n = 18) stated that they learned something about SE by playing the Simulation Game. Interestingly the more experienced students of Group 2 perceived occurred learning even stronger than those of Group 1.86% of Group 2 chose the answer item representing the highest agreement. 89.5% (n = 17) of all participants stated they would play the Simulation Game again and students would recommend the game to a friend or fellow student. Students stated they would play the game again to optimize their performance and to improve their high score. They would like to benefit from experiences of the first run and to test findings in a next one. Students stated that learning was perceived faster and deeper as in lectures. Asked for the reasons, why they recommend the Simulation Game as a standard part of SE courses, participants stated, that they appreciated the focus on essential content, the relatively short duration of the game, the overview, and the preparation of upcoming projects provided. The students liked that relationships between elements, for example, Activities and Alphas, got visible. Students mentioned fast feedback, focus on virtually practicing steps in an endeavor and the opportunity to learn from failures as particularly advantageous. Students perceived to have improved their performance while playing the Simulation Game. This seems to be a good indicator of felt selfefficacy and fits well to the perceived learning. The majority, 79% of the participants (n = 15), stated that they reinforced SE knowledge of previous lectures or courses. Students stated too that they perceived that they learned something new about SE by playing the Simulation Game. They
Developing Essential Software Engineering Competencies Supported
83
stated that having to think about the next steps to take action and to have to repeatedly process Alphas and their Checkpoints, to recurrently think about them in detail, helped to recognize relationships between elements and to internalize concepts. All (100%, n = 19) students stated that the fun at playing the game, the necessity to make own active decisions, and the Essence Navigator helped at learning. Of these features, active decision-making, was assessed as very strongly helping by the highest number of participants (79%, n = 15| Group 1: 75%, n = 9|Group 2: 86%, n = 6) followed by fun (74%, n = 14|Group 1: 75%, n = 9|Group 2: 71%, n = 5). These results indicate strong arguments for DGBL and inherent constructivist learning approaches. That the Essence Navigator was perceived as that helpful, confirms the assumption that digital learning games should embed learning content about the process, method, or kernel in this case. This was an aspect identified as lacking by most of the existing DGBL approaches so far. Students stated that debriefing activities (89%, n = 17), as well as team communication (84%, n = 16), provided learning. Out of them, almost one-third of the students indicated these two aspects as very strongly supporting learning, and two-thirds considered them as helpful. The author observed that the game provided triggers and initiated social interactions. Active and direct face-to-face communication was preferred over utilizing the elementary chat provided by the game. As one student stated: “We did not need the team chat because we were sitting next to each other.” These results indicate strong arguments for social constructivist learning approaches. Individual team results strongly indicate that providing the opportunity to play as a team may work well, but does not guarantee such desired cooperation. Depending on the initial situation of teams’ members, the short period allocated to playing the game in the case study may not have provided enough opportunity to get to that state. The time travel feature could not prove its support of learning with statistical significance (Sign-test, α = 0.05) It seems that not all students recognized this feature as supportive as anticipated. 63% (n = 12) of participating students stated that this feature was helpful at learning. One-half of them perceived this feature as very strongly helping. This assertion is supported by the free text answers stating several times that students appreciated the “opportunity to learn from mistakes.” Results show that only 63% of all players used this feature. This value of 63% matches the ratio of players appreciating the time travel feature as supportive of learning. Since the answers to the questionnaire were collected anonymously, no clear correlation can be determined, but the results may indicate that players using this feature appreciate it as supportive at learning. Other features built into the Simulation Game that were not able to show statistically significant support (Sign-test, α = 0.05) to learning are rankings and leaderboards. Only 42% (n = 8) of participating students stated that rankings and leaderboards provided support at learning, 58% of the students (n = 11) did not. As students’ agreement to helping team communication indicates, the game provided triggers for interaction. Why rankings and leaderboards were not perceived supportive by the majority of players might have several reasons and be caused by the dynamics of the gameplay of the teams. Once team communication got started and had been ongoing, not all team members had to look at the leaderboards provided. Results may have been compared as part of verbal team communication too. Such assumption is
84
Software Engineering Perspectives in Computer Game Development
supported by the observation that ten out of the eleven players not valuing rankings and leaderboards assessed team communication as supportive to learning. Responding students stated that they think SEMAT Essence and the Essence Kernel will be of help in their future projects. None of the students reported that they believe it would not provide support. Students stated that they want to deploy SEMAT Essence in their future projects. While none of the students stated not to want to use Essence in future projects, a third (n = 6) of all participating students chose the neutral item at answering this question.
3.6.4 Course Project Work Students of Group 1 utilized the Essence Navigator, provided in addition to a given SE process, to drive their course project work. As part of regular mandatory team meetings, hold two times a week, the Alpha States and their Checkpoints were assessed. To assess the Alpha States in real project required to think about the project’s progress and health. Different from the Simulation Game, teams now had to provide feedback on their own–triggered by Checkpoints of the Alpha States, which guided group discussions. Students were able to utilize the already familiar Essence Navigator mostly right from the start of the project work and without further support by the lecturer. This indicates that concepts were understood, and students were able to apply them. At some points and depending on individual working style, some teams had slightly more a tendency to quickly answer questions and assess the Checkpoints in a rather superficial way than others. Based on limited experience developing more complex software systems, this comes as no real surprise and may reflect the approach of any novice acting in SE projects. By asking to explain the reasoning behind the assessment of the Checkpoints and the Alpha States, those teams were motivated to think further about specific issues. Altogether it can be summarized that the working style of the teams in the case study was much more oriented toward a holistic perspective of the SE endeavor than the working style of teams in the years before, where the same course was held. Teams discussed much more about essential aspects that got ignored in the past where only the SE process was given to the teams. Thereby the teams did not lose the focus on operative SE tasks to do in the process but were additionally able to give reasons for doing those tasks. Three out of four teams performed the optional mapping exercise and delivered reasonable results. As students perceived the activities to do, besides designing and implementing source code, as very comprehensive, they were surprised their assigned project work did not cover some essential aspects of an SE endeavor for organizational reasons of the course. Those aspects got clearly visible by the non-addressed Alpha States and put the project work done into perspective.
3.7 DISCUSSION AND CONCLUSIONS Corresponding to the findings of other deployments of the Essence Kernel [52], some ambiguous formulations and single terms used in the descriptions of checkpoints provided reasons for requesting and discussions. Péraire and Sedano reported that
Developing Essential Software Engineering Competencies Supported
85
these ambiguities were leading to situations “where the team discusses the meaning of a checklist item instead of having a conversation about the project” [24]. In the context of the described course project, aiming at introducing students into SE practices and methods, such discussions were not judged as impeding but welcome since they provided occasions to reflect on approaches prescribed by the process provided and to think about alternatives. Since students of Group 1 mainly had no chance to develop more complex software systems in the past, they were not able to draw on any existing experiences utilizing SE practices covering the essential aspects of an SE endeavor. That is why they were provided with an SE process tailored to the need of the course. Since the given SE process was not implemented using the Essence language, some concepts had to be mapped and discussed. For instance, the given SE process made use of Use Cases to structure requirements. The Essence Kernel on its own does not provide substructures of requirements since different practices handle this in various ways. This challenging aspect was reported by Péraire and Sedano [52] too. Differing from their study students in this case study did not already have considerable average work experience in industrial practice. The observation that students were able to utilize the Essence Navigator, as well as the provided enacted Essence Kernel, right from the start of their course project work indicates that the Simulation Game contributed to lowering the cognitive load that students faced in their project work. Following the Cognitive Load Theory (CLT), any knowledge already learned, in terms of CLT stored in long-term memory, frees room in the rather limited working memory and facilitates learning of related new knowledge. [53] This indicates too that the Simulation Game provided a valuable anchor that students were able to refer to in their thinking and discussions with teammates. Introducing additional concepts like Essence takes time. Alpha assessment sessions as part of regular team meetings need time too. Without time added to a course, the time available for remaining tasks is inevitably reduced. Students of Group 1 in the case study were able to deliver their projects with at least the same quality as teams on the same course in years before. Some of the less prioritized features of the required software project outcome had to be omitted in the course project in favor of a more holistic and conscious working style. By utilizing the generalist approach of the SEMAT Essence Kernel, students were provided with a highly transferable thinking framework acting as an anchor and providing support, orientation, and guidance in any future SE endeavor. All in all, students applied a systematic, disciplined and quantifiable approach to the development of their software—and appreciated the guidance given by that approach. Hence, they acted as software engineers-to-be. The results indicate that the decision to develop games with a focus on the intended learning objectives as well as the design decisions introduced in Section 4 vitally contributed to the performance of the approaches taken. The case study, conducted to evaluate the developed games, provided first insights and encouraging results. Since the number of participants was rather small, results cannot be considered to be conclusive. A wider utilization of the proposed approaches is needed to compare the results with those of the case study and to collect a broader base for evaluation.
86
Software Engineering Perspectives in Computer Game Development
3.8 FUTURE WORK AND RESEARCH DIRECTIONS Since the conduction of the case study described in this chapter, much activity inside the SEMAT community and their global SEMAT Education Project created valuable learning material, including a textbook and accompanying material like exercises and quizzes. The games introduced in this chapter will get integrated into the efforts to make SEMAT Essence easily accessible to the broad community in education and training. To accomplish that, the Simulation Game needs to be adapted to learning scenarios different than the in-class and face-to-face courses that it was designed for initially. Team dynamics while playing the Simulation Game provide an exciting field of research for future work. The game elements, which were not able to show statistically significant support at learning, provide interesting fields of observation and research for future studies of the game. It would be interesting to prove or disprove the hypothesis that players, assessing time travels as well as rankings and leaderboards as not supportive to learning, draw primarily from team’s experience and team communication—and may be trying to avoid failures already made by others. To support such findings, the results of the game and the questionnaires would need some mapping—something that was not implemented in this first study. The broad inclusion of Essence in SE education worldwide opens opportunities for replicated studies to evaluate the effectiveness and efficiency of the approach further. The learning material that is now available allows for experimental study designs including comparison groups.
ACKNOWLEDGMENTS Most findings presented in this chapter are the results of the author’s Ph.D. research, which would not have been possible without the support of the author’s university and faculty. In particular, the author would like to thank the supervisors and reviewers of his dissertation for stimulating discussions. Special thanks go to the IACS for the opportunities of exchanging ideas and supporting the presentation of author’s ideas and results at colloquia, workshops, and conferences. Essential impulses for author’s work resulted from these events.
NOTES 1 In this chapter, the term SE method is preferred to the term software process. Both are used in literature and community partly synonymously, but also in parts with different semantics. SWEBOK states “[...] methods provide an organized and systematic approach to developing software [...].” Computer Society [1] Depending on the context, a software process or SE method may represent a complete life cycle (SDLC or SPLC), but it does not have to necessarily. 2 following the categorization of Prensky [7]. 3 Software Engineering Method and Theory, an initiative that “drives a process to re-found software engineering based on a solid theory, proven principles and best practices.” [8], http://semat.org 4 The PDCA cycle represents an adapted Demen or Shewart cycle [11], well known in the field of modern quality control and continuous improvement and, as such, part of lean thinking principles.
Developing Essential Software Engineering Competencies Supported
87
5 This work uses the widely accepted Revised Bloom’s Taxonomy [42] to characterize learning objectives. 6 This characterization utilizes those proposed by Prensky [7]. 7 https://puzzler.sim4seed.org
REFERENCES [1] IEEE Computer Society. 2014. Guide to the Software Engineering Body of Knowledge (SWEBOK(R)): Version 3.0 (3rd ed.). IEEE Computer Society Press, Los Alamitos, CA, USA. [2] T. C. Lethbridge, J. Diaz-Herrera, R. J. LeBlanc, and J. Barrie Thompson. 2007. Improving Software Practice Through Education: Challenges and Future Trends. In 2007 Future of Software Engineering (FOSE’07). IEEE Computer Society, Washington, DC, USA, 12–28. doi: 10.1109/FOSE.2007.13 [3] B. Boehm. 2006. A View of 20th and 21st Century Software Engineering. In Proceedings of the 28th International Conference on Software Engineering (ICSE’06). ACM, New York, NY, USA, 12–29. doi: 10.1145/1134285.1134288 [4] I. Sommerville. 2010. Software Engineering (9th revised ed.). Addison-Wesley Longman, Amsterdam. [5] C. G. von Wangenheim and F. Shull. 2009. To Game or Not to Game? IEEE Software 26, 2, 92–94. [6] K. Reich. 2012. Konstruktivistische Didaktik: Das Lehr- und Studienbuch mit OnlineMethodenpool (5. erweiterte auflage ed.). Beltz. [7] M. Prensky. 2007. Digital Game-Based Learning. Paragon House. [8] SEMAT Community. 2018. What is SEMAT? http://semat.org/what-is-semat [9] Object Management Group (OMG). 2015. Essence - Kernel and Language for Software Engineering Methods, Version 1.1. http://www.omg.org/spec/Essence/1.1/ [10] The SEMAT Community. 2016. SEMAT - Alpha State Cards. http://semat.org/de/ alpha-state-cards-with-abbrev-checklists [11] R. Moen and C. Norman. 2006. Evolution of the PDCA cycle. http://pkpinc.com/files/ NA01_Moen_Norman_fullpaper.pdf [12] I. Jacobson, P.-W. Ng, P. E. McMahon, I. Spence, and S. Lidman. 2013. The Essence of Software Engineering: Applying the SEMAT Kernel (1st ed.). Addison-Wesley Professional, Upper Saddle River, NJ. [13] P. E. McMahon. 2013. Essence: Why do we need it? http://sematblog.wordpress. com/2013/11/16/essence-why-do-we-need-it/ [14] I. Jacobson, P.-W. Ng, I. Spence, and P. E. McMahon. 2014. Major-league SEMAT: Why should an executive care? Commun. ACM 57, 4, 44–50. http://dl.acm.org/citation. cfm?id=2580712 [15] I. Jacobson, I. Spence, and P.-W. Ng. 2013. Agile and SEMAT - Perfect Partners. Queue 11, 9, 30:30–30:41. doi: 10.1145/2538031.2541674 [16] P.-W. Ng and S. Huang. 2013. Essence: A framework to help bridge the gap between software engineering education and industry needs. IEEE, 304–308. doi: 10.1109/ CSEET.2013.6595266 [17] M. Thomas Connolly, E. A. Boyle, E. MacArthur, T. Hainey, and J. M. Boyle. 2012. A systematic literature review of empirical evidence on computer games and serious games. Computers & Education 59, 2 (Sept. 2012), 661–686. doi: 10.1016/j. compedu.2012.03.004 [18] S. Jiang, H. Zhang, C. Gao, D. Shao, and G. Rong. 2015. Process Simulation for Software Engineering Education. In Proceedings of the 2015 International Conference on Software and System Process (ICSSP 2015). ACM, New York, NY, USA, 147–156. doi: 10.1145/2785592.2785606
88
Software Engineering Perspectives in Computer Game Development
[19] D. C. C. Peixoto, R. M. Possa, R. F. Resende, and C. I. P. S. Pádua. 2011. An Overview of the Main Design Characteristics of Simulation Games in Software Engineering Education. In 2011 24th IEEE-CS Conference on Software Engineering Education and Training (CSEE T). 101–110. https://doi.org/10.1109/CSEET.2011.5876076 [20] A. Nassal. 2015. Projektmanagement Spielend Lernen. In Software Engineering im Unterricht der Hochschulen (SEUH). 53–64. [21] J. Pieper. 2017. Simulation and Digital Game-based Learning in Software Engineering Education: An Integrated Approach to Learn Software Engineering Methods. Ph.D. Dissertation. Universität Rostock, Rostock. [22] E. Navarro. 2006. SimSE: A Software Engineering Simulation Environment for Software Process Education. Ph.D. Dissertation. University of California, Irvine, CA. http://www. ics.uci.edu/~emilyo/papers/Dissertation.pdf [23] E. O. Navarro and A. van der Hoek. 2004. SIMSE: An Interactive Simulation Game for Software Engineering Education. In Proceedings of the 7th IASTED International Conference on Computers and Advanced Technology in Education (CATE). Kauai, Hawaii, 12–17. [24] K. Shaw and J. Dermoudy. 2005. Engendering an Empathy for Software Engineering. In Proceedings of the 7th Australasian conference on Computing education - Volume 42 (ACE ’05). Australian Computer Society, Inc., Darlinghurst, Australia, 135–144. http:// dl.acm.org/citation.cfm?id=1082424.1082441 [25] A Drappa and J. Ludewig. 1999. Quantitative modeling for the interactive simulation of software projects. Journal of Systems and Software 46, 2–3, 113–122. doi: 10.1016/ S0164-1212(99)00005-9 [26] A. Drappa and J. Ludewig. 2000. Simulation in software engineering training. In Proceedings of the 22nd International Conference on Software Engineering (ICSE’00). ACM, New York, NY, USA, 199–208. doi: 10.1145/337180.337203 [27] P. Mandl-Striegnitz. 2001. How to Successfully Use Software Project Simulation for Educating Software Project Managers. In Frontiers in Education Conference, 2001. 31st Annual, Vol. 1. IEEE, T2D–19. http://ieeexplore.ieee.org/xpls/abs_all. jsp?arnumber=963884 [28] R. T. Mittermeir, E. Hochmüller, A. Bollin, S. Jäger, and M. Nusser. 2003. AMEISE – A Media Education Initiative for Software Engineering Concepts, the Environment and Initial Experiences. In Proceedings of the Interactive Computer aided Learning (ICL) 2003 International Workshop. Villach, Austria. [29] M. O. Barros, A. R. Dantas, G. O. Veronese, and C. M. L. Werner. 2006. Model-driven game development: Experience and model enhancements in software project management education. Software Process: Improvement and Practice 11, 4, 411–421. [30] H. Sharp and P. Hall. 2000. An Interactive Multimedia Software House Simulation for Postgraduate Software Engineers. In Software Engineering, International Conference on. IEEE Computer Society, Los Alamitos, CA, USA, 688. doi: 10.1109/ ICSE.2000.10053 [31] E. Ye, C. Liu, and J. A. Polack-Wahl. 2007. Enhancing Software Engineering Education Using Teaching Aids In 3-D Online Virtual Worlds. In 2007 37th Annual Frontiers In Education Conference-Global Engineering: Knowledge Without Borders, Opportunities Without Passports. IEEE, T1E–8. http://ieeexplore.ieee.org/xpls/abs_all. jsp?arnumber=4417884 [32] A. Jain and B. Boehm. 2006. SimVBSE: Developing a game for value-based software engineering. In Proceedings of the 19th Conference on Software Engineering Education and Training. 103–114.
Developing Essential Software Engineering Competencies Supported
89
[33] C. Caulfield. 2011. Shall We Play a Game? Ph.D. Dissertation. Edith Cowan University, Joondalup, Australia. http://ro.ecu.edu.au/theses/447/ [34] A. Nassal. 2014. A General Framework For Software Project Management Simulation Games. In 2014 9th Iberian Conference on Information Systems and Technologies (CISTI). 1–5. doi: 10.1109/CISTI.2014.6877074 [35] A. Nassal and M. Tichy. 2016. Modeling Human Behavior for Software Engineering Simulation Games. In Proceedings of the 5th International Workshop on Games and Software Engineering (GAS ’16). ACM, New York, NY, USA, 8–14. doi: 10.1145/ 2896958.2896961 [36] D. C. C. Peixoto, R. M. Possa, R. F. Resende, and C. I. P. S. Pádua. 2012. Challenges and Issues in the Development of a Software Engineering Simulation Game. In 2012 Frontiers in Education Conference Proceedings. 1–6. doi: 10.1109/FIE.2012.6462318 [37] J. E. N. Lino, M. A. Paludo, F. V. Binder, S. Reinehr, and A. Malucelli. 2015. Project Management Game 2d (Pmg-2d): A Serious Game To Assist Software Project Managers Training. In IEEE Frontiers in Education Conference (FIE), 2015.32614 2015.1–8. doi: 10.1109/FIE.2015.7344168 [38] P. A. Kirschner, J. Sweller, and R. E. Clark. 2006. Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problembased, experiential, and inquiry-based teaching. Educational Psychologist 41, 2, 75–86. [39] L. Harasim. 2011. Learning Theory and Online Technologies. Routledge, New York, NY. [40] Fred Paas and John Sweller. 2014. Implications of Cognitive Load Theory for Multimedia Learning. In R. Mayer (Ed.), The Cambridge Handbook of Multimedia Learning, Cambridge University Press, Cambridge, UK (Cambridge Handbooks in Psychology, pp. 27–42). [41] C. M. Reigeluth. 1999. The Elaboration Theory: Guidance for Scope and Sequence Decisions. In Instructional Design Theories and Models: A New Paradigm of Instructional Theory. Vol. 2. Lawrence Erlbaum Associates, Mahwah, NJ, 425–453. https://books.google.de/books?hl=de&lr=&id=FW9BA3c_VRkC&oi=fnd&pg =PT438&dq=The+elaboration+theory:+Guidance+for+Scope+and+Sequences+Decisi ons&ots=hq8AVzgEQQ&sig=CKlb4oNo3GK9G7NbLIKw9GecpDU [42] L. W. Anderson and D. Krathwohl. 2001. A Taxonomy for Learning, Teaching, and Assessing: A Revision of Bloom’s Taxonomy of Educational Objectives. Addison Wesley, Longman, New York. [43] J. S. Breuer and G. Bente. 2010. Why so serious? On the relation of serious games and learning. Eludamos. Journal for Computer Game Culture 4, 1, 7–24. [44] R. Van Eck. 2006. Digital game-based learning: It’s not just the digital natives who are restless. Educause Review 41, 2, 16–30. [45] C. Dondi and M. Moretti. 2007. A methodological proposal for learning games selection and quality assessment. British Journal of Educational Technology 38, 3, 502–512. doi: 10.1111/j.1467-8535.2007.00713.x [46] D. R. Krathwohl. 2002. A revision of Bloom’s taxonomy: An overview. Theory into Practice 41, 4, 212–218. http://www.tandfonline.com/doi/pdf/10.1207/ s15430421tip4104_2 [47] B. W. Tuckman and M. A. C. Jensen. 1977. Stages of small-group development revisited. Group & Organization Management 2, 4, 419–427. http://gom.sagepub.com/content/2/4/419.short [48] S. de Freitas and F. Liarokapis. 2011. Serious Games: A New Paradigm for Education? In Serious Games and Edutainment Applications. Springer, 9–23. http://link.springer. com/chapter/10.1007/978-1-4471-2161-9_2
90
Software Engineering Perspectives in Computer Game Development
[49] U. Ritterfeld and R. Weber. 2006. Video games for entertainment and education. In Playing Video Games: Motives, Responses, and Consequences, P. Vorderer and J. Bryant (Eds.). Lawrence Erlbaum Associates, Mahwah, NJ, 399–413. http://ocw.metu.edu.tr/ pluginfile.php/2382/mod_resource/content/0/ceit706/week7/Ritterfeld_Weber.PDF [50] D. H. Jonassen. 1994. Thinking Technology: Toward a Constructivist Design Model. Educational Technology 34, 4, 34–37. [51] V. J. Shute, M. Ventura, M. Bauer, and D. Zapata-Rivera. 2009. Melding the power of serious games and embedded assessment to monitor and foster learning. Serious Games: Mechanisms and Effects 2 (2009), 295–321. https://books.google.de/books?hl=de&lr= &id=eGORAgAAQBAJ&oi=fnd&pg=PA295&dq=Melding+the+Power+of+Serious+ Games+and+Embedded+Assessment+to+Monitor+and+Foster+Learning:+Flow+and+ Grow.&ots=1u8qv2LcRv&sig=geOS7ZfG-DA-rpWbzod0uQCbwJQ [52] C. Péraire and T. Sedano. 2014. State-Based Monitoring and Goal-Driven Project Steering: Field Study of the Semat Essence Framework. In Companion Proceedings of the 36th International Conference on Software Engineering (ICSE Companion 2014). ACM, New York, NY, USA, 325–334. doi: 10.1145/2591062.2591155 [53] J. J. G. van Merriënboer and J. Sweller. 2005. Cognitive Load Theory and Complex Learning: Recent Developments and Future Directions. Educational Psychology Review 17, 2, 147–177. doi: 10.1007/s10648-005-3951-0
4
Building Casual Games and APIs for Teaching Introductory Programming Concepts Brian Chau, Rob Nash, and Kelvin Sung Comp. & Software Sys., University of Washington Bothell
Jason Pace Digital Future Lab, University of Washington Bothell
CONTENTS 4.1 Introduction����������������������������������������������������������������������������������������������������� 91 4.2 Games and CS1/2 Classes������������������������������������������������������������������������������� 92 4.3 API Design Guidelines������������������������������������������������������������������������������������ 93 4.4 The Development Processes���������������������������������������������������������������������������� 94 4.4.1 The Two Games����������������������������������������������������������������������������������94 4.4.2 The Teams������������������������������������������������������������������������������������������� 95 4.4.3 Game Development Process���������������������������������������������������������������� 95 4.4.4 API Definition and Refinement Process���������������������������������������������� 96 4.5 Results������������������������������������������������������������������������������������������������������������� 96 4.5.1 The Space Smasher API���������������������������������������������������������������������� 96 4.5.2 The Corrupted API������������������������������������������������������������������������������ 98 4.5.3 Discussions������������������������������������������������������������������������������������������ 99 4.6 Conclusion������������������������������������������������������������������������������������������������������99 Acknowledgments�������������������������������������������������������������������������������������������������� 100 Notes���������������������������������������������������������������������������������������������������������������������� 100 References�������������������������������������������������������������������������������������������������������������� 100
4.1 INTRODUCTION When properly integrated into coursework, using videogames to teach computer science (CS) accomplishes desired student learning outcomes, builds excitement and enthusiasm for the discipline, and attracts a bright new generation of students early in their academic careers [1, 2]. As a relatively new approach, interested faculty require assistance in the form of elementary sample materials and tutorials to support their exploration and experimentation [3]. 91
92
Software Engineering Perspectives in Computer Game Development
The Game-Themed Computer Science (GTCS) project and the associated library [3] are designed specifically for this purpose with elaborate sets of sample teaching materials that hide the graphics and gaming details [4]. The self-contained nature of the materials allows faculty to adopt and use each without significant modification to their existing classes. Results from adopting GTCS materials have demonstrated effectiveness in engaging students and achieving the desired learning outcomes [5– 7]. In addition, results from the many workshops (e.g., [8]) showed that although interested faculty members with no background in graphics or gaming found the GTCS materials to be non-trivial, they were able to comprehend and begin developing game applications based on the GTCS library within a matter of hours [3]. Student feedback on GTCS materials indicated that though they find the materials motivating, they were also frustrated by the simplicity, for example, the absence of fundamental gaming features like power-ups or win conditions. On the other hand, faculty workshop participants pointed out that the most demanding efforts in building game-specific teaching materials are often unrelated to the educational goals which include the time-intensive processes of locating or generating art and audio assets, or implementing the annoying details of various object interaction rules. To address this seemingly contradictory feedback while preserving the important characteristics of simplicity and usability for a targeted curriculum, the GTCS project group is building a series of causal games and corresponding APIs. These standalone games each showcase one or two programming concepts, allowing faculty to pick and choose for selective adoption. The games have gone through elaborate playtesting to ensure an engaging and complete gameplay experience. Each API is methodically extracted from the finished game and refined based on usability and support for the presentation of targeted programming concepts so that faculty can build their own aesthetically engaging materials while focusing on the pedagogy rather than irrelevant details such as asset management. Currently, there are five games under development in various stages of completion. Two of the games in particular—Space Smasher and Corrupted—include finalized APIs, and the sample teaching materials for Space Smasher are currently being field tested in CS1/2 classrooms. This paper uses these two games and their respective APIs as examples to discuss our game development and API refinement processes and results. In the rest of this paper, Section 2 briefly surveys previous work; section 3 reviews existing API studies and articulates a design guideline for our game APIs; section 4 discusses our game and API development processes; section 5 presents our APIs from Space Smasher and Corrupted; and section 6 concludes the paper.
4.2 GAMES AND CS1/2 CLASSES Existing work on presenting CS1/2 concepts in the context of computer games can be broadly categorized into three approaches [4]: little or no game programming (e.g., [9]) where students learn by playing custom games; per-assignment game development (e.g., [10]) where individual assignments are games designed around technical topics being studied; and extensive game development where faculty and students work with custom game engines (e.g., [11]) or specialized environments (e.g., [12]).
Building Casual Games and APIs
93
As pointed out by Levy and Ben-Ari [13] and Ni [14], issues that faculty consider when examining new and innovative teaching materials for adoption include (among others) preparation time, material contents, and topic coverage. Yet, most of the existing results from integrating games in CS1/2 classes are typically from faculty members with expertise in graphics or games and are “student-centric,” where the main goals are student engagement and various learning outcomes—preparation time for adoption and flexibility of the materials for topic coverage are usually not primary concerns. Indeed, it can be challenging to take advantage of these results for the general faculty teaching CS1/2, since many have little to no background in computer graphics or games. As discussed in the previous section, while effective in addressing the issues of faculty background and curricula modularity, the feedback from previous GTCS materials identified the seemingly contradictory student desire for complexity and faculty need for simplicity [3]. This paper presents the recent GTCS project group efforts in addressing this interesting dichotomy—building complete casual games that offer meaningful gameplay experiences for the students while designing APIs that ensure simple and straightforward curriculum development for the faculty.
4.3 API DESIGN GUIDELINES An API can be described as a well-defined interface that exposes the external services of a singular component to clients who will consume these services as elemental software building blocks [15]. In our case of designing an interface for developing casual games, our APIs are a collection of functions intended to be reused by other programmers to perform common tasks that may be difficult, cumbersome, or tedious [16]. Well-defined APIs foster productivity, code reuse, encapsulation of complex systems, and consistent behaviors for their users [17]. A sound API should be easy to use and hard to misuse [18]. In the design of our API, some of the fundamental goals overlap (e.g., productivity and code reuse) while others oppose one another, offering interesting challenges. For example, while achieving tight encapsulation is desirable, our APIs must purposefully expose details relating to the concepts that they are designed to teach, for example, exposing the details of underlying 2D arrays that represent the grid system of Corrupted for student manipulation. To add to this, our end users are on the two ends of the spectrum of programming expertise—faculty and students of CS1/2 classes. To address our requirements, the following considerations are articulated to guide the design of our APIs. Usability and Structured Simplicity: usability and simplicity facilitate the creation of effective CS1/2 materials accessible to students and faculty with no background in graphics or gaming Discoverability and Learnability: our end users must be able to build simple applications quickly with minimum familiarity of the APIs but also have the opportunity to gradually explore advanced functionality at their own pace
94
Software Engineering Perspectives in Computer Game Development
Expressiveness and Productivity: while the final products built by the students may resemble casual games, the APIs primarily support the building of effective teaching materials as vehicles for educational content delivery Encapsulation and Modularity: as previously discussed, the goals are to tightly encapsulate the complex graphics and gaming functionality while strategically exposing selected concepts for teaching purposes. Though important, performance is only a referencing factor in our API design—as long as an acceptable frame rate and memory footprint are maintained.
4.4 THE DEVELOPMENT PROCESSES The ideal game design and implementation for our purposes must be simple enough so that neither students nor faculty become mired in graphics or game complexity, and yet the gameplay must be genuinely fun so that students can connect their work to a final experience that stands on its own merits. Fortunately, simple and fun are not mutually exclusive when it comes to game design.
4.4.1 The Two Games Space Smasher is a variant of Super Breakout-style games, where players remove blocks on the screen by bouncing a ball with a moveable paddle. The popularity of this type of game spans generations, and the numerous variants tend to be largely identical and feature basic color block graphics. Space Smasher introduces more interesting gameplay by adding customizable blocks that are capable of triggering events such as swapping blocks or enabling/disabling unbreakable blocks. Also included are more premium sets of graphics tiles and sounds than are typically found in this genre. The ball-block-paddle collision tests, the special event logic, and the iteration through all blocks present an excellent structured sandbox for teaching and playing with conditionals and loops conceptually. Corrupted is a variation on the Bubble Shooter genre, where players launch a colored tile into a larger group of tiles and attempt to make matches of three until all tiles are removed or until the tiles advance to meet the player at the opposite end of the screen [19]. Bubble shooters also tend to be simple match-and-remove games with minimalist graphics. Corrupted recasts the game with an active automated opponent employing a variety of tricks to increase the challenge and intrigue, and the game itself has been given a distinct artistic style. The visual, spatial, and multidimensional aspects of the color tiles present a rich domain of concepts for use in teaching 1D and 2D arrays. Note that it is relatively straightforward to design custom levels and additional gameplay elements to both Space Smasher and Corrupted. The game mechanics for either do not require extensive balancing or tuning to make levels enjoyable. Even novice designers can quickly create fun and challenging custom levels. In this way, after a basic process of discovery and familiarization with the concepts and APIs, students can implement their own unique levels as practice exercises.
Building Casual Games and APIs
95
4.4.2 The Teams We knew from the outset that creating casual games designed both to teach fundamental programming concepts and to engage players would require teams with varying expertise: making fun games is an interdisciplinary undertaking and requires a wide variety of skills. Thus, we established a partnership that includes frontline CS1/2 faculty members and the Digital Future Lab (DFL)1 of the University of Washington Bothell (UWB) campus—an interactive media R&D studio developing original interactive works supporting education, entertainment, and social justice. The DFL emphasizes students’ dual role as creators and learners, while working with students across disciplines and backgrounds to ensure that each core component of the game is well designed. The games are developed almost entirely by undergraduate students, including CS and non-CS majors. Roles include level design, visual and user-interface design, sound and music composition, game development, testing, and project management.
4.4.3 Game Development Process Our design approach begins with a unique brainstorming process where game designers generate ideas using familiar game mechanics to facilitate learning while practicing a specific programming concept, and also to gauge the fun factor of the prescribed activities. Faculty contributors help guide the design toward modular and feasible outcomes by providing simple and clearly defined coding requirements. The photo in Figure 4.1 exemplifies our unique approach with the initial whiteboard sketches for the Space Smasher game. Notice that the left side of the board charts the “if” control structure that the game is being designed to teach, the center shows a gameplay screen mockup, and basic game rules are listed on the right. The conditional programming construct (on the left) provides the underlying impulse guiding the design process from its initial stages. As an initial idea gains momentum, the DFL designers examine and refine fundamental gaming mechanics to maximize overall entertainment value, while CS1/2 faculty evaluate implementation, simplicity, and the level of exposure to the intended programming concept. Simultaneously, student developers create simple digital prototypes so designers can experiment interactively with their ideas and make
FIGURE 4.1 Initial whiteboard sketches for Space Smasher.
96
Software Engineering Perspectives in Computer Game Development
improvements to the core gaming experience. Playable versions of the game then undergo rigorous hands-on testing to refine each design choice or mechanic.
4.4.4 API Definition and Refinement Process The final game and corresponding API development overlap significantly through a three-step process: (1) finalize the game prototype, (2) define and refine the API while completing the game implementation, and (3) build tutorials and teaching materials based on the API. In our process, the API refinement spans from the stabilization of the prototype game until the team finishes tutorials and teaching materials. This allows for verification of initial usability of the APIs [20]. Once these pieces are in place, the team fine-tunes gameplay while integrating production–quality graphic and audio assets. The result is a game that has the lookand-feel of a studio-quality production, while containing library and game features that will challenge new CS students to program individual game variations as part of a larger learning process.
4.5 RESULTS The APIs are defined based on our previous experience from the GTCS foundations game engine [6] where the user code subclasses from an API-defined superclass and overrides two protected methods: initialize() and update(). The API calls the initialize() method exactly once before the game begins and the update() method continuously at a real-time rate until the game ends. The underlying philosophy of the API is to provide all the functionality such that user code can focus on implementing just the game logic that targets the selected CS1/2 concepts. This adds responsibilities to the API as it must anticipate and provide a slew of resources to accomplish such as exercises, including pre-defined win and lose menu screens, access to all art and audio assets (user code can override these if desired), and anything related to the gameplay environment (e.g., window size and UI layout). In addition, the APIs provide access to each in-game object (e.g., balls, paddles, or tiles) and their behaviors (e.g., move, speed, and remove) plus all potential object-to-object interactions (e.g., ball-and-block collisions and reflections). All created objects are drawn automatically unless explicitly removed or set to invisible. To ensure user testability, mouse and keyboard input are both supported and special debugging modes are built into each API (e.g., stopping and allowing the player to control the ball movement in Space Smasher or key binding to create specific-colored tiles in Corrupted). In the following we present actual sample teaching materials for Space Smasher and Corrupted to further illustrate each API.
4.5.1 The Space Smasher API The screen shots (left: full game, right: teaching example) and code listing in Figure 4.2 show an example for teaching conditionals and loops. Note that the MySpaceSmasherGame class is a subclass of the API defined SpaceSmasher
97
Building Casual Games and APIs
public class MySpaceSmasherGame extends SpaceSmasher { // init the game, called once by the API protected void initialize() { lifeSet.add(5); //create 5 lives: show in top-left paddleSet.add(1); //create 1 paddle blockSet.setBlocksPerRow(6); //num of blocks per row for (int i = 0; i < 2; i++) { //create two rows of blocks blockSet.addNormalBlock(1); //normal block (light gray) blockSet.addFireBlock(1); //fire block (red) blockSet.addNormalBlock(1); //normal block (light gray) blockSet.addFireBlock(1); //fire block (red) blockSet.addFreezingBlock(1); //ice block (blue) blockSet.addFreezingBlock(1); //ice block (blue) } } //update is called continuously >40 times per second protected void update() { //control the paddle left/right movement Paddle paddle = paddleSet.get(0); if (keyboard.isButtonDown(KeyEvent.VK_LEFT)) paddle.moveLeft(); if (keyboard.isButtonDown(KeyEvent.VK_RIGHT)) paddle.moveRight();
}
}
//get the paddle //move paddle left //move paddle right
//conditionally spawning balls with loops if (keyboard.isButtonDown(KeyEvent.VK_1)) { Ball foo = new Ball(); //make a new ball ballSet.add(foo); //add to set of balls foo.spawn(paddle); //put it on screen above the paddle } else if (keyboard.isButtonDown(KeyEvent.VK_2)) { for (int i=0; i α ≥ 0.7 are acceptable, between 0.9 > α ≥ 0.8 are good, and α ≥ 0.9 are excellent [36]. We analyzed the Cronbach’s alpha coefficient separately (Table 5.5) for the sample with data collected from the evaluations of digital games (n = 187), with data collected from non-digital games (n = 402) and both type of games (n = 589), only discarding the items specific to the evaluation of digital games in the latter two (items 10, 11, and 12). Analyzing the 35 standardized items of the MEEGA+ measurement instrument, the value of Cronbach’s alpha is excellent considering the data from digital games (α = .914) as well as non-digital games (α = .939). And, grouping the data from digital and non-digital games (excluding only the items to evaluate digital games), the value of Cronbach’s alpha is again considered excellent (α = .929). Therefore, we can conclude that the answers to the items are consistent and precise, indicating the reliability of the standardized items of the MEEGA+ measurement instrument. Evaluation of the Customized Items Analyzing the reliability of the items that are customized in accordance with the specific learning objectives of each educational game, we again, obtained an excellent Cronbach’s alpha coefficient (α = .953). This also indicates an excellent reliability of this part of the MEEGA+ measurement instrument 5.5.3.2 Construct Validity AQ2: How do underlying factors influence the responses on the items of the MEEGA+ measurement instrument? In order to analyze the conceptual structure in terms of the number of quality factors that represents the responses of the 35 standardized items of the MEEGA+ measurement instrument, we performed an exploratory factor analysis. In order to analyze whether the items of the MEEGA+ measurement instrument can be submitted to the factor analysis process [78], we used the Kaiser-Meyer-Olkin (KMO) index and Bartlett’s test of sphericity being the most commonly used ones [78]. These methods indicate how much the realization of the exploratory factor analysis is appropriate for a specific set of items [78]. The KMO index measures the sampling adequacy with values between 0 and 1. An index value near 1.0 supports a factor analysis and anything less than 0.5 is probably not amenable to useful factor analysis [79]. Bartlett’s sphericity test also indicates whether the factor analysis is appropriate, considering acceptable values of a significance level